Psychological Influences on Surgical Recovery

Perspectives From Psychoneuroimmunology

Janice K. Kiecolt-Glaser, Gayle G. Page, Phillip T. Marucha, Robert C. MacCallum, and Ronald Glaser
The Ohio State University

Greater fear or distress prior to surgery is associated immune and endocrine function (Liebeskind, 1991).
with a slower and more complicated postoperative recov- Thus, viewing the psychological literature on postsurgical
ery. Although anxiety presumably interferes with recuper- recovery within a PNI context suggests a new conceptual
ation through both behavioral and physiological mecha- framework, illustrated in Figure 1.
nisms, the pathways have been unclear. Recent work in The model suggests that psychological variables
psychoneuroimmunology (PNI) has demonstrated that could influence wound healing, a key variable in short-
stress delays wound healing. In addition, a second line term postsurgical recovery, through several pathways: (a)
of research has illustrated the adverse effects of pain on Emotions have direct effects on "stress" hormones, and
endocrine and immune function. A biobehavioral model they, in turn, can modulate immune function. (b) The
is described that is based on these and other data; it patient's emotional response to surgery can influence the
suggests a number of routes through which psychological type and amount of anesthetic, and anesthetics vary in
and behavioral responses can influence surgery and post- their effects on the immune and endocrine systems. (c)
surgical outcomes. Clinical and research implications Certain health behaviors may dictate differences in
are highlighted. choice of anesthetic (e.g., alcohol intake), or extent of
surgery (e.g., obesity). In addition, health habits such as
smoking that are themselves stress-responsive can have
direct deleterious consequences for immune function and

S urgery is a threatening experience, with multi-

ple stressful components--concerns about one's
physical condition, admission to a hospital, antici-
pation of painful procedures, worries about survival and
recovery, and separation from family. Accordingly, it is
wound healing. (d) Individuals who are more anxious are
also more likely to experience greater postsurgical pain,
and pain can down-regulate immune function. Thus, im-
mune function, already poorer as a consequence of pre-
not surprising that even operations that physicians con- surgical stress, could decline even further.
sider " m i n o r " can provoke strong emotional reactions After providing a brief overview of key events in
in patients (Johnston, 1988). If these psychological re- wound healing, we discuss the model's components and
sponses are sufficiently intense, they can have important paths. Literature addressing psychological influences on
consequences: The weight of the evidence suggests that recovery from surgery is viewed within the context of
greater distress or anxiety prior to surgery is associated our framework. Underscoring fundamental methodologi-
with a slower and more complicated postoperative recov- cal concerns that emerge as a consequence of our concep-
ery (Johnston & Wallace, 1990; Mathews & Ridgeway, tualization, we suggest that researchers need to consider
1981). Moreover, researchers who have assessed the im- each of the pathways to maximize the signal-to-noise
pact of psychosocial interventions administered before ratio. We end by highlighting clinical issues and
surgery have generally demonstrated positive effects on applications.
postsurgical psychological and physical function (see re-
views by Contrada, Leventhal, & Anderson, 1994; De- Editor's note. Cynthia D. Belar served as action editor for this article.
vine, 1992; Johnston & V'6gele, 1993; Johnston & Wal- Author's note. Janice K. Kiecolt-Glaser, Department of Psychiatry,
lace, !990; Mumford, Schlesinger, & Glass, 1982; Suls & College of Medicine, The Ohio State University; Gayle G. Page, Depart-
Wan, 1989). ment of Community, Parent-Child, and Psychiatric Nursing, College
Although anxiety presumably interferes with recov- of Nursing, The Ohio State University; Phillip T. Marucha, Department
ery through both behavioral and physiological mecha- of Periodontology, College of Dentistry, The Ohio State University;
Robert C. MacCallum, Department of Psychology, College of Social
nisms, the pathways for such effects have been unclear. and Behavioral Sciences, The Ohio State University; Ronald Glaser,
Recent work in psychoneuroimmunology (PNI) has pro- Department of Medical Microbiology and Immunology, College of
vided evidence that stress delays wound repair (Kiecolt- Medicine, The Ohio State University.
Glaser, Marucha, Malarkey, Mercado, & Glaser, 1995; Work on this article was supported in part by National Institutes
of Health Grants K02 MH01467 and R37 MH42096.
Marucha, Kiecolt-Glaser, & Favagehi, 1998; Padgett,
Correspondence concerning this article should be addressed to
Marucha, & Sheridan, 1998). In addition, a second line Janice K. Kiecolt-Glaser, Department of Psychiatry, Ohio State Univer-
of research has illustrated the adverse effects of pain on sity College of Medicine, 1670 Upham Drive, Columbus, OH 43210.

November 1998 • American Psychologist 1209

Copyright1998by the AmericanPsychologicalAssociation,Inc. 0003-066X/98/$2.00
Vol.53, No. 11, 1209-1218
Figure 1
A Biobehavioral Mode! of the Psychological, Behavioral, and Biological Pathways That Influence Wound
Healing: The Central Short-Term Outcome in Recovery From Surgery

HEALTH

/I
nAVE I | It I ",. "~
SURGERY I | I
| ? I " - ~ TYPE/EXTENT I • IMMUNE WOUND
*'
PSYCHOLOGICAIJ
I/
IsuOFERY
SURGERY [FUNCTION IHEAUNG
STATUS I \ ~ - - ~ "---'J f / f

NEUROENDOCRINE
FUNCTION

Note. This figure is intended to illustrate our conceptual framework for understanding the complex relationships among relevant domains rather than providing a
formal testable model. For simplicity, most paths are described as moving in one causal direction, and all possible connections are not illustrated (e.g., psychological
variables such as stress and appraisal may influence the relationship between extent of surgery and immune function). Psychological appraisals of the resulting
physical stressor may modulate immune function and wound healing. In addition, direct connections between the central nervous system (CNS) and immune function
are also possible, through direct CNS innervation of lymphoid tissue (Fehen & Felten, 1991 ).

Wound Healing: Immune and the ability of fibroblasts and epithelial cells to remodel
Neuroendocrine Influences the damaged tissue (Lowry, 1993).
A number of hormones and neuropeptides can in-
Wound repair progresses through several overlapping fluence this immunological cascade (Htibner et al., 1996;
stages (Htibner et al., 1996). In the initial inflammatory Zwilling, 1994). For example, mice treated with gluco-
stage, vasoconstriction and blood coagulation are fol- corticoids showed impairments in the induction of IL-1
lowed by platelet activation and the release of platelet- and TNF, as well as defects in wound repair (Htibner et
derived growth factors (Van De Kerkhof, Van Bergen, al., 1996). Human studies have also demonstrated that
Spruijt, & Kuiper, 1994). The factors act as chemoattrac-
stress-induced elevations in glucocorticoids can tran-
tants for the migration of phagocytes (neutrophils and
siently suppress both IL-1 and TNF production (DeRijk
monocytes) to the site, starting the proliferative phase
et al., 1997). Accordingly, dysregulation of glucocorti-
that involves the recruitment and replication of cells nec-
coid secretion provides one obvious mechanism by
essary for tissue regeneration and capillary regrowth.
Wound remodeling, the final step, may continue for which stress can alter wound healing.
weeks or months. The healing process is a cascade, and In contrast to the generally negative effects of gluco-
success in the later stages of wound repair is highly de- corticoids, growth hormone (GH) can enhance wound
pendent on initial events (HUbner et al., 1996). healing (Veldhuis & Iranmanesh, 1996). GH serves as a
Immune function plays a key role early in this cas- macrophage activator (Zwilling, 1994); among its activi-
cade. Proinflammatory cytokines such as interleukin 1 ties, it stimulates monocyte migration. GH also amplifes
(IL-1) and tumor necrosis factor (TNF) are essential to superoxide anion generation and bacterial killing by mac-
this effort; they help to protect a person from infection rophages, influential mechanisms for protection from in-
and prepare injured tissue for repair by enhancing phago- fection after wounding (Zwilling, 1994).
cytic cell recruitment and activation (Lowry, 1993). Fur- Although acute stressors can provoke transient in-
thermore, cytokines released by recruited cells regulate creases in GH (Kiecolt-Glaser, Malarkey, Cacioppo, &

1210 November 1998 • American Psychologist

Glaser, 1994), deep sleep is the normal stimulus for that disruption o f neuroendocrine homeostasis modulates
release of much of our daily GH release (Veldhuis & wound healing.
Iranmanesh, 1996); thus,, stressors that modify the Further research assessed the impact of a brief com-
architecture of sleep can also lessen GH secretion. monplace stressor, academic examinations, on alterations
Moreover, aging decreases GH secretion, and this steep in mucosal wound repair. Prior work has demonstrated
decline may be further accelerated by a chronic stres- that medical students' cellular immune responses during
sor (Veldhuis & Iranmanesh, 1996; Malarkey et al., examinations were poorer than those measured in the
1996). Although only a few of the paths have been same individuals during lower stress periods (Kiecolt-
highlighted here, it is obvious that stress-related alter- Glaser & Glaser, 1991); thus, we were interested in
ations in the hypothalamic-pituitary-adrenal (HPA) whether the distress associated with exams would modu-
axis can have important repercussions for the wound- late mucosal wound healing among a group of 11 dental
healing cascade. students (Marucha et al., 1998). Wounds placed three
days before a major test healed an average of 40% more
Stress and Wound Healing slowly than those made during summer vacation, and
Considerable data from animal and human studies have the differences were quite reliable: No student healed as
demonstrated that stress can alter multiple aspects of im- rapidly during this stressful period as during vacation,
mune function (Glaser & Kiecolt-Glaser, 1994). For ex- and no student produced as much IL-1. Although this
ample, family members who provide care for a relative study demonstrated differences in mucosal wounds, the
with Alzheimer's disease are typically more distressed critical early events in the wound healing process are
than well-matched controls; a number of studies suggest virtually identical for oral and dermal wounds (Wikesj6,
that caregivers also have poorer immune function (Kie- Nilveus, & Selvig, 1992). These data confirmed and ex-
colt-Glaser et al., 1994; Kiecolt-Glaser, Glaser, Grav- tended our prior findings in several important ways.
enstein, Malarkey, & Sheridan, 1996). Consistent with Perhaps the most notable finding was the demonstra-
these differences in immune function, further work tion of how a relatively mild stressor impacted wound repair.
showed that caregivers took an average of 9 days or Like most professional students, these dental students were
24% longer than controls to completely heal a small, experts at taking testsmthey had long histories of per-
standardized wound. Importantly, photographic data forming well under these very conditions. The fact that
demonstrated that the largest differences in wound size something as transient, predictable, and relatively benign as
were most apparent early in the process. examination stress can have significant consequences for
Analyses of the production of IL-1/3 in response to wound healing suggests that other everyday stressors may
lipopolysaccharide (LPS) stimulation showed that care- produce similar deficits in wound repair.
givers' peripheral blood leukocytes (PBLs) produced less It is clear that each of the three studies that assessed
than those of controls (Kiecolt-Glaser et al., 1995). IL- the repair of standardized wounds found marked stress-
1/3 produced early after tissue injury can regulate the related reductions in healing, with delays from 24% to 40%.
production, release, and activation of metalloproteinases, Figure 2 shows effect sizes for these studies, using a simple
which are important in the destruction and remodeling correlation as a measure of the relationship between stress
of the wound; IL-1 also regulates fibroblast chemotaxis and two key outcome variables, healing time and production
and production of collagen (Lowry, 1993). Moreover, IL- of IL-1. Unquestionably, the effects are not just statistically
l stimulates production of other cytokines important for significant, but large in a substantive sense. Of interest is
wound healing, including IL-2, IL-6, and IL-8 (Lowry, the fact that the effect sizes for IL-1 and wound heating
1993). Accordingly, these IL-1/3 data provide evidence parallel each other within the two human studies, consistent
of one immunological mechanism that may underlie care- with the importance of early events in the immunological
givers' deficits in wound repair. cascade for subsequent healing (as described earlier, greater
A subsequent study confirmed and extended the data stress is associated with lower IL-1 levels, but correlations
obtained with caregivers. Mice subjected to restraint are shown as positive to simplify the figure). In addition,
stress healed a punch biopsy wound an average of 27% the effect sizes for wound healing were largest and almost
more slowly than unstressed mice (Padgett et al., 1998). identical in the two paradigms that afforded the greatest
In addition, analysis of the dermal and epidermal layers control over extraneous influences, that is, the Padgett et
surrounding wounds revealed less cellular infiltration in al. (1998) study in which mice had been randomly assigned
restraint-stressed mice than controls one to three days to treatment groups, and the within-subject design whereby
after wounding. To test the hypothesis that the delayed each student served as his or her own control (Marucha et
wound healing in restrained mice reflected HPA activa- al., 1998). These stress-related deficits in wound repair have
tion, Padgett et al. (1998) assessed serum corticosterone; broad implications for surgical recovery.
levels in the stressed group averaged 162.5 ng/ml, com-
pared with 35.7 ng/ml in the controls. Blocking gluco- Pain: Influences on Neuroendocrine and
corticoid receptors in restraint-stressed animals with Immune Function
RU40555 resulted in healing rates comparabl e to control The impact of painful stress on neuroendocrine and im-
animals. Accordingly, these data provided good evidence mune function has been studied both in laboratory ani-

November 1998 • American Psychologist 1211

 The above mentioned findings cannot confirm the
role of pain per se in these neuroendocrine and immune
consequences of painful stress; however, there is evidence
to support such a suggestion. First, several studies have
shown that anesthetic techniques that block transmission
of nociceptive impulses either locally (Pasqualucci et al.,
1994) or at the level of the spinal cord (Koltun et al.,
1996) also significantly reduce neuroendocrine or im-
mune responses affected by surgery. Moreover, when Pas-
qualucci et al. (1994) assessed visual analog scale (VAS)
pain scores, they found that the local anesthetic infiltra-
tion group also exhibited significantly lower pain scores
as well. Second, two prospective studies found that, com-
pared with inhalational anesthesia, epidural anesthesia
was associated with a significant reduction in the inci-
dence of postoperative infections, a biologically signifi-
cant immune outcome (Cuschieri, Morran, Howie, &
McArdle, 1985; Yeager, Glass, Neff, & Brinck-Johnsen,
1987). Cuschieri et al. (1985) also documented signifi-
cantly lower postoperative VAS pain scores in the epi-
dural group. Third, narcotic anesthesia has been shown
to suppress the hormonal response to surgery (Lacou-
menta et al., 1987). Fourth, effective postoperative pain
management with systemic opioids has been associated
with reductions in plasma cortisol levels (Moller,
Dinesen, Sondergard, Knigge, & Kehlet, 1988). Finally,
it was recently shown that providing analgesic doses of
morphine in rats significantly attenuates surgery-induced
increases in the metastasis of a tumor cell line that is
controlled by natural killer cells, providing further evi-
mals and humans. Animal models of painful stress in- dence of the potential negative immune consequences of
clude acute stimuli that do not cause tissue damage such pain with implications for the whole organism (Page,
as footshock and tailshock (e.g., Pezzone, Dohanics, & Ben-Eliyahu, & Liebeskind, 1994; Page, McDonald, &
Rabin, 1994). Such painful stressors have been shown to Ben-Eliyahu, in press).
suppress natural killer cell activity (NKCA), lymphocyte The inflammatory response is an important mecha-
proliferative responses to mitogens, specific antibody nism to consider for its contributions to pain, immunity,
production, and mixed lymphocyte reaction (Liebeskind, and the initiation of wound healing. Tissue damage from
1991). Several reports of stress-induced immune suppres- surgical procedures causes the local release of factors
sion that also incorporated neuroendocrine measures including substance P, bradykinin, prostaglandins, and
demonstrated elevated plasma beta-endorphin (Sacer- histamine. Cytokines such as IL-1 are released from in-
dote, Manfredi, Bianchi, & Panerai, 1994) and corticoste- digenous tissue cells and early recruited cells, such as the
rone levels (Pezzone et al., 1994). neutrophil (Htibner et al., 1996). Together, these factors
The neuroendocrine and immune consequences of initiate the inflammatory response marked by vasodila-
tissue-damaging stimuli such as surgery have been well tion, increased capillary permeability, and the sensitiza-
documented in both animals and humans. Postoperative tion of peripheral afferent nerve fibers. In addition to its
elevations in plasma levels of epinephrine, cortisol, and immune-related roles, IL-1 contributes to local hyperal-
beta-endorphin reflect sympathetic nervous system (SNS) gesia (Schweizer, Feige, Fontana, Muller, & Dinarello,
and HPA axis activation (e.g., Salomaki, Leppaluoto, Lai- 1988), as well as a systemic hyperalgesic state and illness
tinen, Vuolteenaho, & Nuutinen, 1993). Immune suppres- symptoms such as fever and malaise (Watkins et al.,
sion during surgery is evidenced by suppression of 1995). The sensitization of nociceptive fibers manifests
NKCA (e.g., Pollock, Lotzova, & Stanford, 1991), lym- as a decrease in the threshold necessary to initiate noci-
phocyte proliferative responses to mitogens, and changes ceptive impulse transmission, resulting in hyperalgesia
in lymphocyte populations (Tonnessen, Brinklov, Chris- (Woolf, 1994). Prostaglandin E participates in the local
tensen, Olesen, & Madsen, 1987). Further supporting the sensitization of nociceptive fibers (Martin, Basbaum,
connections among the SNS, the HPA axis, and the im- Kwiat, Goetzl, & Levine, 1987), as well as in the suppres-
mune system, coincidental surgery-induced changes in sion of NKCA (Leung, 1989). Faist et al. (1990) showed
these systems also have been demonstrated (Koltun et that inhibiting prostaglandin synthesis with indomethacin
al., 1996). treatment reversed surgery-induced depression of postop-

1212 November 1998 • American Psychologist

erative delayed type hypersensitivity responses and lym- fled endocrine changes in response to surgery (Kehlet,
phocyte proliferation and also resulted in a significant 1997). Alcohol also appears to retard wound healing di-
reduction in the incidence of infection. rectly via delays in cell migration and deposition of colla-
It is clear that pain, like other stressors, adversely gen at the wound site (Benveniste & Thut, 1981). Further-
affects immune function; anesthetics that block transmis- more, individuals who are depressed or anxious are more
sion of nociceptive impulses are also associated with likely to self-medicate with alcohol or other drugs, and
better immune function, as well as reductions in postoper- alcohol abuse can potentiate distress (Grunberg & Baum,
ative infections. As will be discussed shortly, more anx- 1985).
ious patients experience greater postsurgical pain (John- Smoking, another surgical risk behavior, diminishes
ston, 1988). Thus, more distressed individuals could ulti- proliferation of 'fibroblasts and macrophages, reduces
mately be doubly disadvantaged: Immune function, blood flow to the skin through vasoconstriction, and can
already poorer as a consequence of presurgical stress, inhibit enzyme systems for oxidative metabolism and ox-
could decline even further as a consequence of enhanced ygen transport (Silverstein, 1992). In addition to demon-
postsurgical pain. strably slower healing, smokers have higher rates of post-
operative infections, perhaps related to the fact that nico-
Health Status and Health Behaviors tine and other toxins in cigarette smoke depress both
Aging is associated with an increased risk for surgery. primary and secondary immune responses by reducing
Immune function declines with age, particularly func- the chemotactic and phagocytic activities of leukocytes
tional aspects of the cellular immune response (Verhoef, (Silverstein, 1992).
1990); these age-related declines are related to infectious Distressed individuals are more likely to experience
complications, one factor in the increased surgical mor- sleep and appetite disturbances. As noted earlier, deep
tality in the elderly (Thomas & Ritchie, 1995). Moreover, sleep is associated with secretion of GH, a hormone that
age and distress appear to interact to promote immune facilitates wound healing, and fragmented sleep results
down-regulation: Older adults show greater immunologi- in reduced GH secretion; even partial sleep loss one night
cal impairments related to stress or depression than results in elevated cortisol levels the next evening (Le-
younger adults (Herbert & Cohen, 1993; Kiecolt-Glaser, proult, Copinschi, Buxton, & Canter, 1997). Postopera-
Glaser, et al., 1996). tive sleep is typically disturbed as a function of the physi-
Some evidence suggests that surgical stress may also ological stress of surgery, pain, opioid use, noise, and
interact with both age and psychological stress to awakenings from monitoring and nursing procedures
heighten risk for older adults. In work from Linn and (Kehlet, 1997). Thus, postsurgical sleep deprivation may
Jensen (1983), older and younger adults did not differ itself be a significant stressor (Johnston, 1988), and dis-
immunologically prior to elective surgery for hernia re- tress may exacerbate or prolong sleep difficulties through
pair; however, the former had significantly lower prolifer- such mechanisms as intensified pain sensitivity.
ative responses to two mitogens five days after the opera-
Certainly, individuals who experience greater pain
tion. In a related study, Linn, Linn, and Jensen (1983)
are likely to modify other health behaviors besides sleep
divided older and younger patients on the basis of their
in response to this stressor. In addition, some health be-
preoperative anxiety; high anxious older patients had
significantly more complications than the other three haviors may in turn influence pain perception, for exam-
groups. Thus, morbidity and mortality following surgery ple, smoking is associated with greater pain tolerance
are already substantially greater among older adults (Lane, Lefebvre, Rose, & Keefe, 1995).
(Thomas & Ritchie, 1995); further suppression of the The assessment of health behaviors is an important
immune response by psychological stress may put older component of PNI studies (Kiecolt-Glaser & Glaser,
adults at even greater risk. 1988). Moreover, the surgical literature provides clear
Key behavioral risk factors for surgery include evidence for the importance of certain health habits for
smoking, alcohol and drug abuse, and nutrition (Kehlet, recovery (Kehlet, 1997). Accordingly, the assessment of
1997). It is important to note that heightened distress is key risk factors (e.g., smoking and alcohol and drug use)
associated with riskier behavior on all of these dimen- would seem to be an important component of psychologi-
sions (Steptoe, Wardle, Pollard, Canaan, & Davies, cal studies addressing surgical recovery--and yet such
1996), and each of these health habits can alter wound data are frequently omitted. For example, few studies
healing through their effects on immune function; addi- mention any assessment of smoking in the analysis of
tional paths for influence include effects on anesthetic outcomes. Similarly, few researchers have systematically
choices and changes in levels of stress hormones. More- discussed relevant indexes of preoperative health status,
over, these behaviors interact with one another; for exam- even though many postsurgical outcomes of interest are
ple, heavy alcohol use is linked to poorer nutrition (Ben- undoubtedly colored by chronic health problems such as
veniste & Thut, 1981). diabetes. Our model suggests that the assessment of
In addition to nutrition, other consequences of pre- health behavior should be an important component of
operative alcohol abuse include alcohol-related immuno- studies addressing behavioral influences on postsurgical
suppression, subclinical cardiac dysfunction, and ampli- morbidity.

November 1998 • American Psychologist 1213

Psychological Influences on Surgical from coronary artery bypass surgery, initial analyses
Recovery showed that optimism was unrelated to severity of dis-
ease or to other relevant medical variables (Scheier et al.,
Conceptualizing surgery as a short-term stressor, consid- 1989). Subsequent analyses controlled for extensiveness
erable research has focused on how patients' emotional of surgery, severity of disease, and a triad of risk factors
responses influence postoperative recovery. In general, (smoking, hypertension, high cholesterol) before as-
high preoperative fear or stress is predictive of a variety sessing the contribution of optimism to various recovery
of poorer outcomes, including greater pain, longer hospi- indexes. Given the rigor of these efforts, the findings
tal stays, more postoperative complications, and poorer are of particular note: Compared with pessimistic men,
treatment compliance (Johnston, 1988; Mathews & optimistic men fared better on perioperative physiological
Ridgeway, 1981). indexes--they began walking faster after surgery, and
Some of the postoperative repercussions of distress rehabilitation staff rated them as showing a more favor-
may be mediated through variables such as anesthetic able physical recovery.
intake; highly anxious patients require more anesthesia Research from another population has provided evi-
than those who are less distressed (Abbott & Abbott, dence of possible mechanisms that may underlie opti-
1995; Gil, 1984; Johnston, 1988; Markland & Hardy, mists' enhanced postoperative recuperation. Among law
1993). As one consequence, endocrinological and immu- students in their first year of study, optimism was associ-
nological changes secondary to anesthesia could be ated with more positive moods, coping, and differences
greater among more fearful individuals. In addition, Ab- in response to stress, and these differences appeared to
bott and Abbott (1995) noted that the side effects of mediate optimists' better immune function during exami-
various anesthetics, muscle relaxants, and narcotics may nations (Segerstrom, Taylor, Kemeny, & Fahey, 1998).
include a number of the postoperative behaviors that are The optimism data are consistent with evidence that
sometimes used as outcome measures (e.g., vomiting, interventions that alter appraisal, coping, and/or mood
nausea, headache, and pain at the incisional site). Higher may also modulate immune and endocrine function,
doses could presumably increase the severity of thereby enhancing surgical recovery (Kiecolt-Glaser et
symptoms. al., 1985; Kiecolt-Glaser & Glaser, 1992; Manyande et
Across a number of studies, greater self-reported al., 1995). In fact, intervention studies provide some of
anxiety and stress are typically related to more se- the best evidence for the role of psychological factors in
vere postoperative pain (Johnston, 1988; Mathews & surgical recovery. In one of the earliest of such studies
Ridgeway, 1981). In addition to direct effects on endo- (Egbert, Battit, Welch, & Barlett, 1964), anesthesiologists
crine and immune function, the greater pain sensitivity paid brief visits to patients the night before surgery to
of more anxious patients may also have further conse- provide information about typical postsurgical physical
quences for recovery because of differences in compli- sensations and to teach them a relaxation technique de-
ance. For example, breathing exercises can reduce the signed to reduce pain. Those patients who received this
risk of pneumonia, and ambulation decreases the risk of additional visit left the hospital an average of 2.7 days
phlebitis and may improve wound healing (Kehlet, 1997). earlier and required roughly half as much morphine as
However, individuals who are more distressed may be individuals receiving routine care. Among the more than
more cautious about following recommendations for 200 treatment studies that have followed Egbert et al.'s
walking, coughing, or deep breathing because of pain (1964) work (see reviews by Contrada et al., 1994; De-
(Mathews & Ridgeway, 1981). vine, 1992; Gil, 1984; Johnston & V6gele, 1993; John-
Linn, Linn, and Klimas (1988) assessed the rele- ston & Wallace, 1990; Mumford et al., 1982; Suls &
vance of differences in preoperative pain tolerance and Wan, 1989), beneficial outcomes have included decreased
stress to postoperative immune function; physiological anxiety and stress, reductions in hospital stay, fewer post-
responses to a cold pressor test were measured the day operative complications, better treatment compliance,
before surgery in 24 men undergoing hernia repair. After less pain and lower use of analgesics, and alterations
controlling for preoperative immunological values (as in physiological indexes (primarily cardiovascular and
well as age and social support), lymphocytes from men respiratory measures). 1 Several studies have suggested
who reported more recent stressful life events had lower that presurgical psychological status can influence physi-
proliferative responses to phytohemagglutinin (PHA). In ological responses during the surgery itself (Abbott &
addition, high responders to the cold pressor stress (i.e., Abbott, 1995; Greene, Zeichner, Roberts, Callahan, &
a lower pain threshold) had significantly lower prolifera- Granados, 1989; Markland & Hardy, 1993; Scheier et al.,
tive responses to pokeweed mitogen after surgery; they 1989) as well as speed of physical recovery (time to open
also required more pain medication and had more
complications.
Personality variables may moderate postsurgical Because most surgical wounds are sutured, direct assessments
outcomes via their influences on stress, mood, and coping of healing are not possible. In addition, although infectionis a highly
(Mathews & Ridgeway, 1981). In an excellent study that relevant complicationwithin our PNI context,it usually occurs at base
assessed the effect of dispositional optimism on recovery rates too low to be useful for research purposes.

1214 November 1998 • American Psychologist

eyes) following the discontinuation of anesthesia (Liu, How do such modest interventions produce such
Barry, & Weinman, 1994). differences? Recall the magnitude of the effect sizes for
A meta-analysis by Johnston and V6gele (1993) ad- stress and wound healing (Figure 2); those data suggest
dressed relationships between the major types of inter- that even treatments that have relatively small conse-
ventions (procedural intervention, sensory information, quences for psychological distress could translate into
behavioral instruction, cognitive intervention, relaxation, faster repair through the direct and indirect routes sug-
hypnosis, or emotion-focused intervention) and eight out- gested by our model.
come variables (negative affect, pain, pain medication, The positive effects of interventions are even more
length of stay, behavioral and clinical indexes of recovery, striking in view of the significant limitations of many
physiological indexes, and satisfaction). Using both pub- studies. These have included heterogeneity in patient
lished and unpublished studies, they concluded that each groups, in surgical and anesthetic procedures, and in the
of the eight outcomes showed significant benefit. Among types of outcomes assessed; failure to assess and control
the interventions, procedural information and behavioral for patients' prior experience with surgery or other inva-
instruction produced the most ubiquitous improvements sive medical procedures; differential treatment of com-
across all eight outcomes, whereas relaxation studies parison groups on variables other than those being manip-
showed benefits for all indexes except behavioral ulated; and the absence of manipulation checks to demon-
recovery. strate that interventions were delivered in a uniform
Summarizing results from several meta-analyses of manner, or that patients actually acquired or utilized new
presurgical intervention studies, Contrada et al. (1994) skills or implemented cognitive or behavioral strategies
argued that the association between surgical preparation (see critiques by Anderson & Masur, 1983; Gil, 1984;
and outcome is clinically meaningful. Depending on the Johnston & V6gele, 1993; Ludwick-Rosenthal & Neu-
meta-analysis, two thirds to three quarters of intervention field, 1988). In addition, we suggest that many of the
patients had better outcomes than controls, and the size published studies may also have underestimated effect
of the improvement was 20% to 28%. Moreover, the sizes because they failed to consider relevant health be-
effect sizes for interventions with two or three content haviors that can make sizable contributions to error
categories (information, coping skills training, and psy- variance.
chosocial support) were larger than those that had only In the same vein, differences in pain and analgesic
a single thrust (Contrada et al., 1994). use should also be assessed. Our model highlights the
In contrast to their efficacy, most presurgical inter- importance of effective pain control, particularly in the
ventions have been remarkably brief. For example, the immediate postoperative period. As discussed earlier, the
psychoeducational treatments included in Devine's healing process is a cascade, with success in the later
stages of wound repair highly dependent on initial events
(1992) meta-analysis took 7 to 90 minutes, with a median
(Htibner et al., 1996); it is important to note that effective
length of 30 minutes; the great majority were limited to
postoperative pain management with systemic opioids
a single session. Moreover, delivery was frequently not
has been associated with reductions in plasma cortisol,
individualized; some studies used a group format,
a hormone strongly associated with poorer wound healing
whereas others relied on such low-cost alternatives as
(Moller et al., 1988). Although most of the researchers
booklets, manuals, audiotapes, and videotapes. Nonethe-
studying psychological influences on postsurgical out-
less, 79% to 84% of the studies showed beneficial effects
comes have conceptualized pain as an outcome, it also
(depending on the outcome) for pain, psychological dis-
contributes to the process: Poorly controlled acute pain
tress, and various indexes of recovery. In fact, 79% of could retard wound healing and prolong the recovery
the studies showed that these interventions were associ- period (Liebeskind, 1991; Page et al., 1994, in press).
ated with a shorter length of hospital stay: Compared
with controls, treatment patients spent an average of 1.5 Implications for Research and Practice
fewer days in the hospital. In the United States, 80% of all surgeries are considered
Indeed, positive effects are not limited to formal "elective" in the sense that the patient can choose when
interventions; one study suggested that even small envi- to have the operation (Sobel & Ornstein, 1996). Early
ronmental differences may be beneficial. Hospital records assessment of psychological predictors of outcome pro-
of gall bladder surgery from t 972 to 1981 were examined vides the opportunity to identify those patients who might
to determine whether assignment to a room with a view be at greater risk (e.g., Block, 1996). However, to provide
from a window made a difference (Ulrich, 1984). Consis- maximally effective interventions, further information
tent with studies that have documented the utility of dis- about the impact of individual differences in coping style
traction or attention redirection as a coping strategy to and level of anticipatory anxiety would be valuable. For
reduce distress and pain (Gil, 1984), comparisons of example, "monitors," patients with a vigilant coping
well-matched patient pairs showed that those with a view style, scan for threat-relevant cues, whereas "blunters"
had shorter hospital stays (7.96 vs. 8.70 days) and that rely more on avoidant coping (Miller, 1992). The two
they took fewer potent analgesics than those without a groups appear to show better adjustment in health-related
view. contexts when interventions are tailored to their coping

November 1998 • American Psychologist 1215

style. Monitors fare better with voluminous sensory and The Kulik et al. (1996) data illustrate an important
procedural information; blunters show the opposite re- theme in this literature: Whatever their content, many
sponse (Miller, 1992; Prokop, Bradley, Burish, Ander- presurgical interventions may benefit recovery in part
son, & Fox, 1991). Such data are particularly important because they provide additional interpersonal support
precisely because there is less opportunity to provide during a stressful time (Contrada et al., 1994; Johnston,
care for patients following surgery (Deardorff & Reeves, 1988). There is certainly ample evidence that social sup-
1997). port can moderate the effects of psychological stress; in
As a consequence of technical advances, many sur- addition, a number of studies have shown relationships
geries are less invasive than in the past and thus are between social support and dimensions of autonomic,
now performed on an outpatient basis, or with a greatly endocrine, and immune function, with family ties ap-
reduced hospital stay (Macho & Gable, 1994). In addi- pearing to be a key source of support related to physio-
tion, managed care guidelines are increasingly con- logical functioning (Kiecolt-Glaser, Newton, et al., 1996;
straining the length of hospitalization. With any extended Uchino, Cacioppo, & Kiecolt-Glaser, 1996). Consistent
convalescence at home, family members play more prom- with this literature, male coronary bypass patients who
inent roles. The self-regulatory model proposed by Con- received greater spousal support used less pain medica-
trada et al. (1994) provides guidance regarding interven- tion, had a more rapid discharge from the surgical inten-
tion content that explicitly addresses the patient, the pa- sive care unit, and spent fewer total days in the hospital
tient's spouse or partner, and the patient-partner (Kulik & Mahler, 1989).
relationship. In this context, the demands that at-home Indeed, one factor that should spur research in this
recuperation may place on family members who provide arena is its obvious and immediate practical applicability.
care should not be overlooked, and neither should the Compared with the costs of hospitalization, psychologi-
potential toll for caregivers (see also Kiecolt-Glaser et cal interventions are clearly cost-effective (Contrada et
al., 1994). al., 1994; Devine, 1992; Mumford et al., 1982).
Our knowledge of how psychological interventions As described earlier, dental students took an average
are actually translated into faster physical recovery is of 40% longer to heal a small, standardized wound made
limited by the range of measures used in most studies. prior to examinations, compared with an identical wound
Despite the fact that reduction of anxiety or stress is a made during vacation (Marucha et al., 1998). In contrast
goal of many of the interventions, the kinds of outcomes to the relatively mild and predictable stress of academic
assessed do not typically include data from each of the examinations, surgery is a high-stakes stressor, with pos-
relevant domains: self-report, behavioral, and physiologi- sible consequences that include death, pain, disfigure-
cal. Although the addition of endocrine and immune mea- ment, economic losses, alterations in social roles, and
sures both pre- and postoperatively would be worthwhile uncertainty about both the outcome and the time course
for psychological studies, appending psychological mea- for recovery (Contrada et al., 1994). Given the multiple
sures to biologically oriented studies would also be quite threats embodied in an approaching surgery, it is not
valuable. For example, researchers who used an indwell- surprising that patients may display marked elevations in
ing catheter to collect blood samples every 20 minutes anxiety for at least six days prior to surgery, with clearly
for 24 hours found that presurgical patients showed re- heightened distress persisting for five to six days after-
markable increases in cortisol secretion as they were be- ward, dropping back to normal levels only after a period
ing prepared preoperatively (body shaving, wash, and of weeks (Johnston, 1980). Stressors that are perceived
enema), with values that spiked 6.9 to 10.5 standard devi- as unpredictable and uncontrollable can continue to be
ations above the mean for control patients for that time associated with elevated stress hormones (Baum, Co-
of day (Czeisler et al., 1976). The interdisciplinary study hen, & Hall, 1993). The ability to "unwind" after stress-
of such phenomena could amplify our understanding of ful encounters, that is, quicker return to one's neuroendo-
psychological and behavioral influences on health. crine baseline, influences the total burden that stressors
Indeed, surgery offers an attractive paradigm for place on an individual (Frankenhaeuser, 1986). Accord-
examining the responses of patients to a distinctive natu- ingly, interventions that promote early adaptation can
ralistic stressor (Johnston, 1988). One innovative study produce substantial benefits for mental and physical
addressed social comparison and affiliation under threat health.
among coronary-bypass patients (Kulik, Mahler, & REFERENCES
Moore, 1996). The men were assigned preoperatively to
Abbott, J., & Abbott, P. (1995). Psychologicaland cardiovascularpre-
a room alone, or they shared a semiprivate room with a dictors of anaesthesia induction,operativeand postoperativecompli-
roommate who was either cardiac or noncardiac, and cations in minor gynecologicalsurgery. British Journal of Clinical
either preoperative or postoperative. Those patients as- Psychology, 34, 613-625.
signed a postoperative roommate were less anxious be- Anderson, K. O., & Masur, F. T., III. (1983). Psychologicalpreparation
fore surgery. In addition, those who either roomed with for invasive medical and dental procedures. Journal of Behavioral
Medicine, 6, 1-40.
a cardiac patient or a postoperative patient walked more Baum, A., Cohen, L., & Hall, M. (1993). Controland intrusivememo-
after their operation, and had shorter hospital stays. The ries as possible determinantsof chronic stress. Psychosomatic Medi-
no-roommate patients had the slowest recoveries. cine, 55, 274-286.

1216 November 1998 • American Psychologist

Benveniste, K., & Thut, P. (1981). The effect of chronic alcoholism on Johnston, M., & Wallace, L. (Eds.). (1990). Stress and medical proce-
wound healing. Proceedings of the Societyfor Experimental Biology dures. Oxford, England: Oxford University Press.
and Medicine, 166, 568-575. Kehlet, H. (1997). Multimodal approach to control postoperative patho-
Block, A.R. (1996). Presurgical psychological screening in chronic physiology and rehabilitation. British Journal of Anaesthesia, 78,
pain syndromes. Hillsdale, NJ: Erlbaum. 606-617.
Contrada, R. J., Leventhal, E. A., & Anderson, J. R. (1994). Psychologi- Kiecolt-Glaser, J.K., & Glaser, R. (1988). Methodological issues in
cal preparation for surgery: Marshaling individual and social re- behavioral immunology research with humans. Brain, Behavior, and
sources to optimize self-regulation. In S. Maes, H. Leventhal, & M. Immunity, 2, 67-78.
Johnson (Eds.), International Review of Health Psychology (Vol. 3, Kiecolt-Glaser, J. K., & Glaser, R. (1991). Stress and immune function
pp. 219-266). New York: Wiley. in humans. In R. Ader, D. Fetten, & N. Cohen (Eds.), Psychoneuroim-
Cuschieri, R. J., Morran, C. G., Howie, J. C., & McArdle, C. S. (1985). munology H (pp. 849-867). San Diego, CA: Academic Press.
Postoperative pain and pulmonary complications: Comparison of Kiecolt-Glaser, J. K., & Glaser, R. (1992). Psychoneuroimmunology:
three analgesic regimens. British Journal of Surger); 72, 495-498. Can psychological interventions modulate immunity? Journal of
Czeisler, C. A., Ede, M. C., Regestein, Q. R., Kisch, E. S., Fang, V. S., & Consulting and Clinical Psychology, 60, 569-575.
Ehrlich, E. N. (1976). Episodic 24-hour cortisol secretory patterns Kiecolt-Glaser, J. K., Glaser, R., Gravenstein, S., Malarkey, W. B., &
in patients awaiting elective cardiac surgery. Journal of Clinical En- Sheridan, J. (1996). Chronic stress alters the immune response to
docrinology and Metabolism, 42, 273-283. influenza virus vaccine in older adults. Proceedings of the National
Deardorff, W.W., & Reeves, J.R. (1997). Preparing for surgery: A Academy of Sciences, 93, 3043-3047.
mind-body approach to enhance healing and recovery. Oakland, Kiecolt-Glaser, J. K., Glaser, R., Williger, D., Stout, J., Messick, G.,
CA: New Harbinger Publications. Sheppard, S., Ricker, D., Romisher, S. C., Briner, W., Bonnell, G., &
DeRijk, R., Michelson, D., Karp, B., Petrides, J., Galliven, E., Deuster, Donnerberg, R. (1985). Psychosocial enhancement of immunocom-
P., Paciotti, G., Gold, P. W., & Sternberg, E. M. (1997). Exercise and petence in a geriatric population. Health Psychology, 4, 25-41.
circadian rhythm-induced variations in plasma cortisol differentially Kiecolt-Glaser, J. K., Malarkey, W., Cacioppo, J. T., & Glaser, R. (1994).
regulate interleukin-l? (IL-I?), IL-6, and tumor necrosis factor-a Stressful personal relationships: Endocrine and immune function. In
(TNF-?) production in humans: High sensitivity of TNF-? and resis- R. Glaser & J. K. Kiecolt-Glaser (Eds.), Handbook of human stress
tance of IL-6. Journal of Clinical Endocrinology and Metabolism, and immunity (pp. 321-339). San Diego, CA: Academic Press.
82, 2182-2192. Kiecolt-Glaser, J. K , Marucha, P.T., Malarkey, W.B., Mercado,
Devine, E. C. (1992). Effects of psychoeducational care for adult surgi- A. M., & Glaser, R. (1995). Slowing of wound healing by psychologi-
cal patients: A meta-analysis of 191 studies. Patient Education and cal stress. Lancet, 346, 1194-1196.
Counseling, 19, 129-142. Kiecolt-Glaser, J. K., Newton, T., Cacioppo, J.T., MacCallum, R. C.,
Egbert, L.D., Battit, G.E., Welch, C.E., & Barlett, M.K. (1964). Glaser, R., & Malarkey, W. B. (1996). Marital conflict and endocrine
Reduction of postoperative pain by encouragement and instruction function: Are men really more physiologically affected than women?
of patients. New England Journal of Medicine, 270, 825-827. Journal of Consulting and Clinical Psychology, 64, 324-332.
Faist, E., Ertel, W., Cohnert, T., Huber, P., Inthom, D., & Heberer, G. Koltun, W. A., Bloomer, M. M., Tilberg, A. E, Seaton, J. E, Ilahi, O.,
(t990). Immunoprotective effects of cyclooxygenase inhibition in Rung, G., Gifford, R. M., & Kauffman, G. L. (1996). Awake epidural
patients with major surgical trauma. Journal of Trauma, 3, 8-18. anesthesia is associated with improved natural killer cell cytotoxicity
Felten, D. L., & Felten, S. Y. (1991). Innervation of lymphoid tissue. and a reduced stress response. American Journal of Surgery, 171,
In R. Ader, D. L. Felten, & N. Cohen (Eds.), Psychoneuroimmunology 68-73.
(pp. 87-101). San Diego, CA: Academic Press. Kulik, J. A , & Mahler, H. L (1989). Social support and recovery from
Frankenhaeuser, M. (1986). A psychobiological framework for research surgery. Health Psychology, 8, 221-238.
on human stress and coping. In M. H. Appley (Ed.), Dynamics of Kulik, J. A., Mahler, H. I., & Moore, P. J. (1996). Social comparison
stress: Physiological, psychological, and social perspectives (pp. and affiliation under threat: Effects on recovery from major surgery.
101-116). New York: Plenum. Journal of Personality and Social Psychology, 7], 967-979.
Gil, K. M. (1984). Coping effectively with invasive medical procedures: Lacoumenta, S., Yeo, T.H., Burrin, J.M., Bloom, S.R., Paterson,
A descriptive model. Clinical Psychology Review, 4, 339-362. J. L., & Hall, G. M. (1987). Fentanyl and the beta-endorphin, ACTH,
Glaser, R., & Kiecolt-Glaser, J. K. (Eds.). (1994). Handbook of human and glucoregulatory hormonal response to surgery. British Journal
stress and immunity. San Diego, CA: Academic Press. of Anaesthesia, 59, 713-720.
Greene, P. G., Zeichner, A., Roberts, N. L., Caltahan, E. J., & Granados, Lane, J. D., Lefebvre, J. C., Rose, J. E , & Keefe, E J. (1995). Effects
J.L. (1989). Preparation for cesarean delivery: A mutticompouent of cigarette smoking on perception of thermal pain. Experimental
analysis of treatment outcome. Journal of Consulting and Clinical and Clinical Psychopharmacology, 3, 140-147.
Psychology, 57, 484-487. Leproult, R., Copinschi, G., Buxton, O., & Cauter, E. V. (1997). Sleep
Grunberg, N.E., & Baum, A. (1985). Biological commonalities of loss results in an elevation of cortisol levels the next evening. Sleep,
stress and substance abuse. In S. Shiffman & T.A. Wills (Eds.), 20, 865-870.
Coping and substance use (pp. 25-62). San Diego, CA: Academic Leung, K. H. (1989). Inhibition of human NK cell and LAK cell cyto-
Press. toxicity and differentiation by PGE2. Cellular Immunology, 123,
Herbert, T. B., & Cohen, S. (1993). Stress and immunity in humans: A 384-395.
meta-analytic review. Psychosomatic Medicine, 55, 364-379. Liebeskind, J. C. (1991). Pain can kill. Pain, 44, 3-4.
Ht~bner, G., Brauchle, M., Smola, H., Madlener, M., Fassler, R., & Linn, B. S., & Jensen, J. (1983). Age and immune response to a surgical
Wemer, S. (1996). Differential regulation of pro-inflammatory cyto- stress. Archives of Surgery, 118, 405-409.
kines during wound healing in normal and glucocorticoid-treated Linn, B. S., Linn, M.W., & Jensen, J. (1983). Surgical stress in the
mice. Cytokine, 8, 548-556. healthy elderly. Journal of the American Geriatric Society, 31, 544-
Johnston, M. (1980). Anxiety in surgical patients. Psychological Medi- 568.
cine, 10, 142-152. Linn, B. S., Linn, M. W., & Klimas, N. G. (1988). Effects of psycho-
Johnston, M. (1988). Impending surgery. In S. Fisher & J. Reason physical stress on surgical outcome. Psychosomatic" Medicine, 50,
(Eds.), Handbook of life stress, cognition and health (pp. 79-100). 230-244.
New York: Wiley. Liu, R., Barry, J. E. S., & Weinman, J. (1994). Effect of background
Johnston, M., & V6gele, C. (1993). Benefits of psychological prepara- stress on postoperative recovery. Anaesthesia, 49, 382-386.
tion for surgery: A meta-analysis. Annals of Behavioral Medicine, Lowry, S. E (1993). Cytokine mediators of immunity and inflammation.
15, 245-256. Archives of Surgery, 28, 1235-1241.

November 1998 • A m e r i c a n P s y c h o l o g i s t 1217

Ludwick-Rosenthal, R., & Neufield, R. W. J. (1988). Stress manage- mune responses and immunocyte beta-endorphin concentrations in
ment during noxious medical procedures: An evaluative review of the rat. Brain, Behavior, and Immunity, 8, 251-260.
outcome studies. Psychological Bulletin, 104, 326-342. Salomaki, T. E., Leppaluoto, J., Laitinen, J. O., Vuolteenaho, O., & Nuu-
Macho, J., & Gable, G. (1994). Everyone's guide to outpatient surgery. tinen, L. S. (1993). Epidural versus intravenous fentanyl for reducing
Toronto, Ontario, Canada: Somerville. hormonal, metabolic, and physiologic responses after thoracotomy.
Malarkey, W. B., Wu, H., Cacioppo, J. T., Malarkey, K. L., Poehlmann, Anesthesiology, 79, 672-679.
K.M., Glaser, R., & Kiecolt-Glaser, J.K. (1996). Chronic stress Scheier, M. E, Matthews, K.A., Owens, J. E, Magovern, G.J., Sr.,
down regulates growth hormone gene expression in peripheral blood Lefebvre, R. C., Abbott, R. A., & Carver, C. S. (1989). Dispositional
mononuclear cells of older adults. Endocrine, 5, 33-39. optimism and recovery from coronary artery bypass surgery: The
Manyande, A., Simon, B., Gettins, D., Stanford, S. C., Mazhero, S., beneficial effects on physical and psychological well-being. Journal
Marks, D. E, & Salmon, P. (1995). Preoperative rehearsal of active of Personality and Social Psychology, 57, 1024-1040.
coping imagery influences subjective and hormonal responses to ab- Schweizer, A., Feige, U., Fontana, A., Muller, K., & Dinarello, C. A.
dominal surgery. Psychosomatic Medicine, 57, 177-182. (1988). Interleukin-1 enhances pain reflexes: Mediation through in-
Markland, D., & Hardy, L. (1993). Anxiety, relaxation, and anaesthesia creased prostaglandin E2 levels. Agents and Actions, 25, 246-251.
for day-case surgery. British Journal of Clinical Psychology, 32, Segerstrom, S. C., Taylor, S. E., Kemeny, M. E., & Fahey, J. L. (1998).
493-504. Optimism is associated with mood, coping, and immune change in
Martin, H. A., Basbaum, A. I., Kwiat, G. C., Goetzl, E. J., & Levine, response to stress. Journal of Personality and Social Psychology,
J.D. (1987). Leukotriene and prostaglandin sensitization of cuta- 74, 1646-1655.
neous high-threshold C- and A-delta mechanonociceptors in the hairy Silverstein, P. (1992). Smoking and wound healing. American Journal
skin of rat hindlimbs. Neuroscience, 22, 651-659. of Medicine, 93, 22S-24S.
Marucha, P. T., Kiecolt-Glaser, J. K., & Favagehi, M. (1998). Mucosal Sobel, D. S., & Ornstein, R. (1996). The healthy mind, healthy body
wound healing is impaired by examination stress. Psychosomatic handbook. New York: Patient Education Media.
Medicine, 60, 362-365. Steptoe, A., Wardle, J., Pollard, T.M., Canaan, L., & Davies, G.J.
Mathews, A., & Ridgeway, V. (1981). Personality and surgical recovery: (1996). Stress, social support and health-related behavior: A study
A review. British Journal of Clinical Psychology, 20, 243-260. of smoking, alcohol consumption and physical exercise. Journal of
Miller, S. (1992). Monitoring and blunting in the face of threat: Implica- Psychosomatic Research, 41, 171 - 180.
tions for adaptation and health. In L. Montada, S. Filipp, & M. J. Suls, J., & Wan, C. K. (1989). Effects of sensory and procedural infor-
Lerner (Eds.), Life crises and experiences of loss in adulthood (pp. mation on coping with stressful medical procedures and pain: A
255-273). Hillsdale, N J: Erlbaum. meta-analysis. Journal of Consulting and Clinical Psychology, 57,
Moiler, I. W., Dinesen, K., Sondergard, S., Knigge, U., & Kehlet, H. 372-379.
(1988). Effect of patient-controlled analgesia on plasma catechol- Thomas, D.R., & Ritchie, C. S. (1995). Preoperative assessment of
amine, cortisol and glucose concentrations after cholecystectomy. older adults. Journal of the American Geriatrics Society, 43, 811-
British Journal of Anaesthesia, 61, 160-164. 821.
Mumford, E., Schlesinger, H. J., & Glass, G. V. (1982). The effect of Tonnessen, E., Brinklov, M. M., Christensen, N.J., Olesen, A. S., &
psychological intervention on recovery from surgery and heart at- Madsen, T. (1987). Natural killer cell activity and lymphocyte func-
tacks: An analysis of the literature. American Journal of Public tion during and after coronary artery bypass grafting in relation to
Health, 72, 141-151. the endocrine stress response. Anesthesiology, 67, 526-533.
Padgett, D. A., Marucha, P. T., & Sheridan, J. E (1998). Restraint stress Uchino, B.N., Cacioppo, J.T., & Kiecolt-Glaser, J.K. (1996). The
slows cutaneous wound healing in mice. Brain, Behavior, & Immu- relationship between social support and physiological processes: A
nity, 12, 64-73. review with emphasis on underlying mechanisms. Psychological Bul-
Page, G. G., Ben-Eliyahu, S., & Liebeskind, J. C, (1994). The role of letin, 119, 488-531.
LGL/NK cells in surgery-induced promotion of metastasis and its Ulrich, R. S. (1984). View from a window may influence recovery from
attenuation by morphine. Brain, Behavior, and Immunity, 8, 241- surgery. Science, 224, 420-421.
250. Van De Kerkhof, P. C. M., Van Bergen, B., Spruijt, K., & Kuiper, J. P.
Page, G. G., McDonald, J. S., & Ben-Eliyahu, S. (in press). Pre- versus (1994). Age-related changes in wound healing. Clinical and Experi-
postoperative administration of morphine: Impact on the neuroendo- mental Dermatology, 19, 369-374.
crine, behavioral, and metastatic-enhancing effects of surgery. British Veldhuis, J. D., & Iranmanesh, A. (1996). Physiological regulation of
Journal of Anaesthesia. the human growth hormone (GH)-insulin-like growth factor type I
Pasqualucci, A., Contardo, R., Da Broi, U., Colt, E, Terrosu, G., Donini, (IGF-I) axis: Predominant impact of age, obesity, gonadal function,
A., Sorrentino, M., Pasetto, A., & Bresadola, E (1994). The effects and sleep. Sleep, 19, $221-$224.
of intraperitoneal local anesthetic on analgesic requirements and en- Verhoef, J. (1990). Transient immunodepression. Journal of Antimicro-
docrine response after laparoscopic cholecystectomy: A randomized bial Chemotherapy, 26, 23-29.
double-blind controlled study. Journal of Laparoendoscopic Surgery, Watkins, L. R., Goehler, L. E., Relton, J. K., Tartaglia, N., Silbert, L.,
4, 405-412. Martin, D., & Maier, S. E (1995). Blockade of interleukin- 1 induced
Pezzone, M.A., Dohanics, J., & Rabin, B.S. (1994). Effects of hyperthermia by subdiaphragmatic vagotomy: Evidence for vagal
footshock stress upon spleen and peripheral blood lymphocyte mito- mediation of immune-brain communication. Neuroscience Letters,
genic responses in rats with lesions of the paraventricular nucleus. 183, 27-31.
Journal of Neuroimmunology, 53, 39-46. Wikesjt, U. M. E., Nilveus, R. E., & Selvig, K. A. (1992). Significance
Pollock, R. E., Lotzova, E., & Stanford, S. D. (1991). Mechanism of of early healing events on periodontal repair: A review. Journal of
surgical stress impairment of human perioperative natural killer cell Periodontology, 63, 158-165.
cytotoxicity. Archives of Surgery, 126, 338-342. Woolf, C. J. (1994). A new strategy for treatment of inflammatory pain:
Prokop, C. K., Bradley, L. A., Burish, T. G., Anderson, K. O., & Fox, Prevention or elimination of central sensitization. Drugs, 47, 1-9.
J. E. (1991). Psychological preparation for stressful medical and den- Yeager, M. P., Glass, D. D., Neff, R. K., & Brinck-Johnsen, T. (1987).
tal procedures. In C. K. Prokop & L. A. Bradley (Eds.), Health psy- Epidural anesthesia and analgesia in high-risk surgical patients. Anes-
chology: Clinical methods and research (pp. 159-196). New York: thesiology, 66, 729-736.
Macmillan. Zwilling, B. S. (1994). Neuroimmunomodulation of macrophage func-
Sacerdote, P., Manfredi, B., Bianchi, M., & Panerai, A. E. (1994). Inter- tion. In J. K. Kiecolt-Glaser & R. Glaser (Eds.), Handbook of human
mittent but not continuous inescapable footshock stress affects im- stress and immunity (pp. 53-76). San Diego, CA: Academic Press.

1218 November 1998 • A m e r i c a n P s y c h o l o g i s t