Content uploaded by Annegret F. Hannawa
Author content
All content in this area was uploaded by Annegret F. Hannawa
Content may be subject to copyright.
Kissing in Marital and Cohabiting
Relationships: Effects on Blood Lipids,
Stress, and Relationship Satisfaction
Kory Floyd, Justin P. Boren, Annegret F. Hannawa,
Colin Hesse, Breanna McEwan, & Alice E. Veksler
Affection exchange theory and previous research suggest that affectionate behavior has
stress-ameliorating effects. On this basis, we hypothesized that increasing affectionate
behavior would effect improvements in physical and psychological conditions known to
be exacerbated by stress. This study tested this proposition by examining the effects of
increased romantic kissing on blood lipids, perceived stress, depression, and relationship
satisfaction. Fifty-two healthy adults who were in marital or cohabiting romantic
relationships provided self-report data for psychological outcomes and blood samples
for hematological tests, and were then randomly assigned to experimental and control
groups for a 6-week trial. Those in the experimental group were instructed to increase
the frequency of romantic kissing in their relationships; those in the control group
received no such instructions. After 6 weeks, psychological and hematological tests were
repeated. Relative to the control group, the experimental group experienced improve-
ments in perceived stress, relationship satisfaction, and total serum cholesterol.
Keywords: Affection; Affection Exchange Theory; Kissing; Lipids
As a nonverbal means of communicating affection, kissing is nearly ubiquitous
among human cultures (Eibl-Eibesfeldt, 1970). Speculations vary as to its origins.
Some suggest that kissing began with the practice of premastication, wherein mothers
Kory Floyd is Professor of human communication at Arizona State University, where the remaining authors are
doctoral students. Order of authorship for all junior authors was determined alphabetically. This research was
supported by grant R03 MH075757-01A1 to the senior author from the National Institute of Mental Health. The
authors are grateful for the technical advice of Jason Short, MD; Harvey Wiener, DO; and Anne Vogel, RN.
Correspondence to: Kory Floyd, Hugh Downs School of Human Communication, Arizona State University,
PO Box 871205, Tempe AZ 85287-1205, USA. E-mail: kory@asu.edu
Western Journal of Communication
Vol. 73, No. 2, April–June 2009, pp. 113–133
ISSN 1057-0314 (print)/ISSN 1745-1027 (online) #2009 Western States Communication Association
DOI: 10.1080/10570310902856071
chewed up food and passed it directly from their lips to their babies’ mouths; this
mouth-to-mouth contact is thought to have then evolved into a more general expres-
sion of care and affection that was later applied to other relationships (Willett, 2001).
Another common idea is that the physical proximity facilitated by kissing allows
for subconscious assessment of the major histocompatibility complex (MHC), an
important predictor of maladaptive mating prospects (Wedekind, Seebeck, Bettins,
& Paepke, 1995). Whatever its origins, there is little question that kissing has a pro-
minent place in the cadre of nonverbal communicative behaviors used to express
interpersonal affection (see Floyd, 2006a; Guerrero & Floyd, 2006) and that it is pro-
vocative biologically, physiologically, and emotionally. To the extent that it mitigates
the psychological and physiological effects of stress, as we will argue herein, kissing
may also effect improvements in physical and mental health conditions that are exacer-
bated by stress. The present experiment was designed to ascertain the effects of
increased romantic kissing on total serum cholesterol, two emotional health parameters
(perceived stress and depression), and assessments of relationship satisfaction.
We begin this essay with a discussion of the biological, physiological, and
emotional components of romantic kissing and a summary of research on its health
outcomes. Next, we advance a theoretic argument, grounded in affection exchange
theory, that romantic kissing within the context of a positively valenced relationship
has stress-ameliorating effects. We then apply the stress-reduction hypothesis to a
hematological outcome, total serum cholesterol, which is known to be exacerbated
by stress. Finally, we apply the hypothesis to emotional health outcomes and
evaluative assessment of the romantic relationship.
Structural Components of Kissing
As a nonverbal form of dyadic affectionate expression in personal relationships,
kissing can be said to involve biological, physiological, and emotional components.
We provide a brief description of each in this section. Although our discussion herein
focuses on the biological, physiological, and emotional components of kissing, we
acknowledge that kissing is also a cultural, symbolic, and sociological behavior, as
are most forms of affectionate expression (see Floyd, 2006a). The parameters of
our discussion reflect the scope and direction of our theoretic argument rather than
a dismissal of these additional components.
Biological Components
At a biological level, kissing is a form of touch. Kissing typically involves direct con-
tact between one partner’s lips and=or tongue and the skin (whether the mouth,
cheek, neck, hand, etc.) or hair of the other partner. Nontactile variations are
observed (e.g., ‘‘blowing a kiss’’), as are nonoral variations (e.g., rubbing noses).
These exceptions aside, however, kissing—whether romantic or platonic—nearly
always entails tactile contact with the lips, and romantic kissing is likely to involve
tactile contact with the lips of both partners simultaneously (Floyd, 2006a). This is
114 K. Floyd et al.
significant from a biological standpoint because, compared to other parts of the
body, the lips are highly sensitive to touch. That is, the lips have both an inordinate
density of dermal and epidermal touch receptors (Stenn & Bhawan, 1990) and large
cortical mapping in the primary somatosensory cortex, where information from
somatosensory nerves is received (Lui & Tang, 2004). Consequently, kissing—parti-
cularly prolonged mutual kissing, involving both sets of lips—stimulates a strong
tactile sensation. Moreover, it is common for kissing to invoke ancillary forms of
touch, such as hugging, hand-holding, and hand-to-face touch, further adding to
the tactile sensory experience (see Floyd, 2006a).
Physiological Components
Romantic kissing likely initiates a range of physiological responses associated with
parasympathetic arousal, salivary exchange, and endocrine reactivity. Although little
research has explicated these effects directly, one can speculate about the specifics of
kissing’s physiological correlates on the basis of research examining other forms of
physical affection. Opining on the general physiological components of positive social
intercourse, for instance, Uvna
¨s-Moberg (1997) wrote that serial interaction within the
context of positive intimate relationships (which frequently involve kissing) ‘‘induces a
psychophysiological response pattern involving sedation, relaxation, decreased sym-
pathoadrenal activity, and increased vagal nerve tone and thereby an endocrine and
metabolic pattern favoring the storage of nutrients and growth’’ (p. 146). In romantic
relationships, physically affectionate interaction in the form of hand holding, cuddling,
and hugging has been shown to have parasympathetic outcomes, including decreased
blood pressure and reductions in the stress hormone cortisol (Grewen, Girdler, Amico,
& Light, 2005; Light, Grewen, & Amico, 2005), as well as elevated epidermal tempera-
ture (Rubinsky, Hoon, Eckerman, & Amberson, 1985; Seeley et al., 1980). It is logical to
expect kissing to be characterized by the same physiological profile.
One of the most socially consequential physiological components of kissing may
involve the exchange of sebum, an oily substance composed of lipids and cellular debris
that is secreted by the sebaceous glands. Sebaceous glands are found in the skin of
mammals, and humans have more than any other mammal (Montagna & Parakkal,
1974). Some researchers (e.g., Comfort, 1974) have suggested that sebum is a semio-
chemical, a chemical that carries a message and affects the behaviors of members of
one’s own species (such as pheromones) or another species (such as allomones).
Nicholson (1984) has argued persuasively that sebum may play a role in feelings of
attachment and bonding, both between sexual partners and between parents and chil-
dren, which would partially explain the ubiquity of kissing as an expression of affection.
Emotional Components
Finally, kissing has an affective component, insofar as it usually occurs within the
context of positively valenced relationships and for the purpose of conveying
affection (Floyd, 2006a). There are exceptions, including the use of kissing in
Western Journal of Communication 115
nonromantic sexual interaction (Leonard, 1990), in perfunctory social greetings
(Greenbaum & Rosenfeld, 1980), and in religious rituals (Penn, 2003). These excep-
tions aside, kissing is generally observed in relationships that are characterized by
positive affect and is exchanged within those relationships as a means of expressing
and reinforcing love, attraction, attachment, and affection, whether romantic or
platonic (Floyd, 2006a). It is therefore logical to expect that kissing increases positive
affect, at least on the part of the kisser, although evidence for this proposition is
primarily anecdotal (e.g., Cane, 2005). Kissing behavior is included, however, in
psychometric indices for relational closeness (Berscheid, Snyder, & Omoto, 1989),
intimacy (Waring, 1984), and affection (Floyd & Morman, 1998), suggesting consen-
sus among relationship researchers that kissing reflects positive affective states.
To the extent that kissing activates biological, physiological, and emotional
processes, it may therefore have implications for health. We review the existing
research subsequently.
Research on the Health Effects of Kissing
Most research on the health effects of kissing has focused on the implications of
salivary exchange and potential blood exchange (via trace amounts of blood in the
saliva) in romantic kissing. These effects include facilitating the transmission of viral
infections, such as influenza (Schoch-Spana, 2000), herpes simplex viruses (Cowan
et al., 2002), and infectious mononucleosis (Carbary, 1975). For instance, a matched
cohort study of adolescents 15 to 19 years of age found that intimate kissing quad-
ruples the risk of meningococcal meningitis (Tully et al., 2006). Romantic kissing
can also facilitate transmission of food allergies (Maloney, Chapman, & Sicherer,
2006) and drug allergies (Liccardi, Gilder, D’Amato, & D’Amato, 2002; Mancuso
& Berdondini, 2006) from one partner to the other. Other scientists have warned
of the potential for avian flu or HIV transmission from romantic kissing if microle-
sions are present on the oral mucosa of the infected partner (Maged, 2006; Piazza
et al., 1989). Still other research has indicated that passionate kissing can change
the DNA composition of saliva, but only within the first minute after kissing has
ceased (Banaschak, Mo
¨ller, & Pfeiffer, 1998).
In addition to its potential negative health outcomes, kissing has also been shown
to elicit some health benefits. Thus far, research on its positive health outcomes has
focused on physiological responses to allergens. In an experiment involving 30
patients with allergic rhinitis (AR) and 30 patients with atopic dermatitis (AD),
for instance, Kimata (2003) found that kissing a romantic partner for 30 minutes sig-
nificantly reduced skin wheal responses to house dust mite and Japanese cedar pollen
(but not histamine), and significantly decreased plasma levels of nerve growth factor,
brain-derived neurotrophic factor, neurotrophin-3, and neurotrophin-4, relative to
nonclinical controls. In a later experiment, Kimata (2006) demonstrated that kissing
for 30 minutes significantly decreased production of allergen-specific immunoglobulin
E(IgE) in atopic patients, relative to nonclinical controls. The findings from these
two studies are important because allergic skin wheal responses and IgE production
116 K. Floyd et al.
are exacerbated by stress in AR and AD patients. Kimata (2003, 2006) reasoned that if
kissing a romantic partner is a stress-alleviating activity, it should therefore precede
significant reductions in these allergic responses. The same logic guides the present
experiment: if romantic kissing ameliorates the psychological and physiological
experience of stress, it should therefore effect improvements in other health outcomes
(besides allergic responses) that are exacerbated by stress.
In the next section, we offer a theoretic and empirical argument for kissing as a
stress-alleviating communicative behavior. Subsequently, we review physiological
and psychological outcomes that have been shown to be aggravated by stress, and
hypothesize that increased kissing can enhance these outcomes.
Kissing as a Stress-Alleviating Behavior
Researchers have long observed that many forms of physical affection behavior in
personal relationships, including kissing, reduce signs of distress. For instance, De
Chateau and Wiberg (1977) reported that when mothers spent more time kissing
their infants at suckling, the infants smiled more and cried less frequently. Affection
exchange theory (AET: Floyd, 2002, 2006a) provides that communicating affection in
close relationships initiates neuroendocrine processes that maximize reward and
buffer the individual against the physiological effects of stress, and that these benefits
are independent of those associated with receiving affectionate expressions. Several
studies have illustrated this pattern. For instance, Floyd (2006b) examined the effects
of expressed and received affection on diurnal variation in the steroid hormone cor-
tisol. Cortisol normally follows a strong diurnal (i.e., 24-hour) rhythm wherein it
peaks immediately after awakening and drops continually during the day, reaching
its lowest point around midnight (Kirschbaum & Hellhammer, 1989). A high degree
of diurnal variation in cortisol levels reflects healthy regulation of the hypothalamic-
pituitary-adrenal axis, one of the body’s principal mechanisms for responding to
acute stress; ‘‘flattened’’ diurnal curves, showing little change in cortisol values from
morning to evening, are therefore indicative of chronic stress (Chrousos & Gold,
1992; Giese, Sephton, Abercrombie, Duran, & Spiegel, 2004; Heim, Ehlert, &
Hellhammer, 2000). As hypothesized, Floyd (2006b) found that, with the effect of
received affection controlled for, expressed affection was directly related to the
magnitude of morning-to-evening change in cortisol (b¼.56).
In a later experiment, Floyd, Mikkelson, Tafoya, et al. (2007) found that during
episodes of acute stress (in which cortisol levels are typically elevated), expressing
affection in writing to a loved one accelerates the return of cortisol to normal levels.
Grewen et al. (2005) similarly found that nonverbal affectionate interaction reduced
cortisol levels for both men and women, and also elevated levels of the neurohypo-
physeal hormone oxytocin in women (see also Turner, Altemus, Enos, Cooper, &
McGuinness, 1999), whereas Floyd, Hesse, and Haynes (2007) found a strong inverse
relationship (b¼.85) between expressed affection and glycohemoglobin (an
index of average blood glucose level, which is elevated by stress), after controlling
for the effects of received affection. In two experiments, Floyd, Mikkelson, Hesse,
Western Journal of Communication 117
and Pauley (2007) also demonstrated that an affectionate writing exercise reduced
total serum cholesterol (which is also elevated by stress) in a group of healthy adults.
The associations between affectionate behavior and well-being are not limited to
physical health outcomes, but extend to mental health as well. Using a compari-
son-groups method, Floyd (2002) demonstrated that highly affectionate adults report
less stress, lower susceptibility to depression, greater overall mental health, and higher
satisfaction with their romantic relationships than do their less affectionate counter-
parts. Given the strongly reciprocal nature of expressed and received affection, one
could conceivably argue that the mental and relational benefits associated with
expressing affection are simply those associated with the amount of affection received
in return. AET provides, however, that communicating affection is beneficial on its
own (i.e., its benefits are independent of those of received affection), and four studies
by Floyd et al. (2005) demonstrated that expressed affection accounts for significant
variance in stress, depression, relationship satisfaction, and mental health even when
received affection is controlled for.
Collectively, these studies reflect AET’s proposition that affectionate behavior
ameliorates the physiological and psychological effects of stress. In the present experi-
ment, we test the ability of romantic kissing to improve a variety of physiological and
psychological outcomes of stress, each of which is detailed subsequently
Physiological Effects of Stress: Total Cholesterol
If affectionate behavior ameliorates stress, as AET predicts and as previous investiga-
tions have established, then it is logical to predict that it will also effect improvements
on physiological parameters that are exacerbated by stress (as Kimata’s experiments
with allergy responses have demonstrated). The present investigation focuses on total
serum cholesterol as a candidate outcome. Cholesterol is a lipid, which is a water-inso-
luble organic compound present in the cell membranes of all body tissues. Cholesterol
performs a number of essential physiological functions, including maintaining
membrane fluidity, producing bile, and contributing to the metabolism of fat-soluble
vitamins (Shier, Butler, & Lewis, 2004). It is also largely responsible for the production
of steroid hormones, such as cortisol, aldosterone, progesterone, the estrogens, and
testosterone. The liver produces most of the body’s cholesterol, although the consump-
tion of foods that are high in cholesterol, trans fat, and=or saturated fat (such as egg
yolks, red meat, full-fat dairy foods, and fried foods) contributes to elevated cholesterol
levels in the bloodstream (Mader, 2005). American Heart Association guidelines pro-
vide that, for healthy adults, total serum cholesterol should be less than 200 mg=dL;
HDL should be above 40 mg=dL for men and above 50 mg=dL for women; LDL should
be less than 100 mg=dL; and triglycerides should be less than 150 mg=dL (American
Heart Association, 2007). Chronically elevated cholesterol (a condition known as
hypercholesterolemia) often leads to the formation and accumulation of plaque deposits
in the arteries, which can contribute to atherosclerosis or coronary heart disease.
Multiple studies have demonstrated that stress is associated with elevations in
total cholesterol and changes in its constituent components, including triglycerides,
118 K. Floyd et al.
high-density lipoproteins (HDL, or ‘‘good cholesterol’’) and low-density lipoproteins
(LDL, or ‘‘bad cholesterol;’’ see, e.g., Bacon, Ring, Lip, & Carroll, 2004; McCann et
al., 1995; Muldoon et al., 1995; Stoney, Niaura, Bausserman, & Metacin, 1999). The
specific mechanisms through which stress elevates cholesterol level are as yet
unknown, although some speculation suggests that they may reflect evolved processes
through which stress-induced increases in energy (in the form of metabolic fuels such
as glucose and fatty acids) initiate ancillary processes that elevate levels of LDL in
the bloodstream (see Steptoe & Brydon, 2005). Other speculation points to activation
of the sympathetic nervous system and the rapid release of catecholamines (such as
epinephrine and norepinephrine) and glucocorticoids (such as cortisol). It has
been documented that lipoprotein lipase activity is inhibited by both norepinephrine
and cortisol (Jansen & Hulsmann, 1985; Miller, Gorski, Oscai, & Palmer, 1989),
which decreases the clearance of triglycerides, decreases HDL concentrations, and
increases LDL concentrations (Huttunen, Ehnolm, Kekki, & Nikkila, 1976).
Because blood lipids are exacerbated by stress, and because affectionate behavior
has stress-ameliorating physiological effects, we therefore hypothesize that increased
kissing improves blood lipids by decreasing total cholesterol. Two experiments by
Floyd, Mikkelson, Hesse, et al. (2007) demonstrated, for instance, that total choles-
terol was decreased as a result of serial affectionate writing; the present trial tests
the ability of a romantic kissing intervention to effect improvements on the same
hematological outcome. Our specific hypothesis was as follows:
H1: In healthy adults, increasing romantic kissing reduces total serum cholesterol.
Psychological Effects of Stress: Depression, Perceived Stress, and Relationship
Satisfaction
We further anticipate that increasing affection in one’s romantic relationship will pro-
duce emotional benefits, including enhanced perceptions of relationship satisfaction
and reductions in perceived stress and depression. Several cross-sectional studies have
established that affectionate behavior in romantic, platonic, and familial relationships is
inversely associated with depression and stress and directly associated with relationship
satisfaction (Floyd, 2002; Floyd et al., 2005; Floyd & Morman, 1998). It is certainly the
case that increases in relationship satisfaction, and=or decreases in stress or depression,
might lead one to become more affectionate—indeed, affectionate behaviors typically
increase in frequency and intensity as romantic love develops (see, e.g., Huston,
Caughlin, Houts, Smith, & George, 2001; Owen, 1987). On the basis of AET, we pro-
pose herein that the obverse causal model may also be operative, wherein increasing
affectionate behavior improves one’s perceptions of satisfaction, stress, and depression.
H2: In healthy adults, increasing romantic kissing reduces perceived stress and
susceptibility to depression.
H3: In healthy adults, increasing romantic kissing increases satisfaction with the
romantic relationship.
Western Journal of Communication 119
Method
Participants
Participants (N¼52) were equal numbers of healthy male and female adults. Ages ran-
gedfrom19to67years,withanaverageof28.63years(SD ¼8.36). Most of the parti-
cipants (78.4%) were Caucasian, whereas 11.8%were Asian=Pacific Islander, 7.8%were
Hispanic, 2.0%were African American, 2.0%were Native American, and 3.9%were of
other ethnic origins (these percentages sum to >100 because participants were allowed
to indicate more than one ethnicity). At the time of the study, 17.6%had completed
some college but had no degree, 2.0%had completed an associate (2-year) degree,
29.4%had completed a baccalaureate (4-year) degree, 49.0%had completed a masters
degree, and 2.0%had completed a professional doctorate (e.g., MD, JD).
Procedures
This study was a federally registered Phase I clinical trial (registry #1001 R03
MH075757-01A1), and was approved by the university’s institutional review board.
Prescreening procedures
Participants were recruited from among the staff, undergraduate student, and grad-
uate student populations at a large university in the southwestern United States. The
study was advertised in three ways: a) via an electronic advertisement on the univer-
sity’s online campus newspaper, b) via flyers posted on bulletin boards around cam-
pus, and c) via an electronic announcement sent to various university listservs. In all
cases, prospective participants were directed to an online prescreening measure to
determine their eligibility for the study. To be considered eligible, prospective parti-
cipants had to: a) be 18 years of age or older; b) be able to speak and read English; c)
be living full-time with a spouse or romantic partner; d) weigh at least 110 pounds; e)
report no history of diagnosis or treatment for hypercholesterolemia; f) report no
current use of blood-thinning agents such as Coumadin; g) report no history of
type 1 or type 2 diabetes; h) report that they were not currently pregnant or breast-
feeding; and i) report no more than mild anxiety about having a capillary blood
draw. A total of 188 prospective participants filled out and submitted the online pre-
screening questionnaire; of that number, 127 (67.6%) met all of the qualifications.
Women and men were equally likely to be qualified for the study (p>.05). The most
common reasons for disqualification were lack of current cohabitation with a roman-
tic partner and body weight of less than 110 pounds.
Laboratory procedures
Qualified participants who consented to take part in the study made an appointment
to visit the Communication Sciences Laboratory and were sent a link to a longer
online questionnaire to fill out beforehand. Participants were instructed to be fasting
when they reported to the lab, having had nothing to eat or drink besides water for at
120 K. Floyd et al.
least 8 hours. Due to the fasting requirement, all lab sessions were scheduled between
7 a.m. and 10 a.m.
When they reported for their Time 1 (T
1
) laboratory visit, participants completed
informed consent forms and were asked about their compliance with the fasting
instructions (all participants reported compliance). A researcher (one of the junior
authors) then activated a Heat Factory (Vista, CA) brand single-use 50C hand
warmer and asked the participant to hold it in his or her nondominant hand while
the participant’s height and weight were recorded. The purpose of the hand warmer
was to stimulate blood flow prior to the capillary puncture. Next, the researcher
retrieved the hand warmer and used a 70%isopropyl alcohol swab to cleanse the
third digit fingertip of the participant’s nondominant hand. Using a 1.75-mm
Tenderlett surgical blade lancet (International Technidyne Corp., Edison, NJ), the
researcher punctured the capillary bed and wiped away the first secretion of blood
with a sterile gauze pad (see McCall & Tankersley, 2003). The researcher then aspi-
rated 80 ml of capillary blood into two glass tubes coated with lithium heparin, an
anticoagulant. Half of the blood was used for the lipid assessment, and the other half
was used for immunocompetence assays not reported here. After the capillary blood
draw, the puncture site was bandaged. The participant was offered juice and a cookie,
was paid $15, and was told to expect an e-mail from the senior author on a specified
date, which would contain additional information and instructions for the study.
Following the 6-week trial (which is described below), participants returned to the
laboratory under the same conditions and instructions for their Time 2 (T
2
) assess-
ments. The only procedural difference between the first and second lab visits was that
participants’ height was only assessed during the first visit; otherwise, the procedures
were identical. Participants were paid an additional $15 at the completion of their
second laboratory visit. For one experimental group participant, the T
2
lipid assess-
ment registered an analytic error after the participant had already left the lab, and we
were not able to reschedule that participant for an additional test; this participant was
therefore excluded from analyses involving changes in total cholesterol.
All researchers involved in the study had received university training in the avoid-
ance of bloodborne pathogens and employed universal precautions while drawing
and handling blood samples, including the use of synthetic (nonlatex) gloves. After
each capillary blood draw, lancets, test materials, and gloves were discarded into
biohazardous waste containers.
Experimental procedures
Assignment to conditions for the 6-week trial was done upon completion of the T
1
laboratory assessments. To ensure an equal sex distribution across conditions, we
used stratified random assignment (via a randomizer software program) to assign
participants to the experimental and control groups. A series of 2 (condition) 2
(sex) ANOVAs confirmed that the experimental and control groups did not differ
significantly from each other in their T
1
values for total cholesterol, relationship
satisfaction, depression, or stress.
Western Journal of Communication 121
Participants in both conditions were e-mailed by the senior author on a
preannounced Monday. All participants were thanked for their participation and
were told to expect an e-mail from the senior author every Monday for the duration
of the 6-week trial. They were told that some of these e-mails would simply contain
information about the study and that some would contain links to short online
questionnaires that participants would be asked to complete.
Participants in the experimental group were also given instructions in the initial
e-mail message that their task during the 6-week trial was to increase, to a noticeable
degree, their frequency and duration of kissing with their cohabiting romantic
partner. The specific text of the message was as follows:
Over the next 6 weeks, we would like you and your spouse or romantic partner to
kiss more frequently than you normally do. At first, you might set aside a few
minutes each day specifically for kissing. Over time, you will probably find that
it becomes a more routine part of how you interact. The point is for the two of
you to kiss each other more often and for longer periods of time than you typically
do right now. We hope you will enjoy this part of the study. It’s fine to tell your
spouse or romantic partner what you have been instructed to do. We hope you will
both make increased kissing a priority over the next 6 weeks.
Participants in both conditions received e-mails from the senior author every
Monday thereafter for the 6-week duration of the trial. The messages on the first,
third, fourth, and sixth Monday consisted of reminders about the study instructions
and indications of when participants would be contacted to schedule their T
2
labora-
tory visits. The messages sent to the experimental group also contained reminders to
continue kissing with elevated frequency. On the second and fifth Mondays, partici-
pants’ messages contained a link to a short online questionnaire that participants
were asked to complete as soon thereafter as possible. The questionnaire contained
items checking the manipulation and indexing changes in participants’ eating and
exercising behaviors and general health.
To minimize experimenter expectancy effects, all of the junior authors (who
conducted the laboratory sessions) were kept blind to the manipulation instructions
and assignment to experimental conditions, and had no access to any of the partici-
pants’ questionnaire data during the study procedure. The senior author, conversely,
was not involved in collecting or analyzing any of the blood samples.
Questionnaire Measures
Relationship satisfaction was measured with the unifactorial 7-item Relationship
Assessment Scale (RAS: Hendrick, 1988). Items included, ‘‘How well does your part-
ner meet your needs?’’ and, ‘‘How good is your relationship, compared to most?’’
Coefficient alpha was .88 at T
1
and .93 at T
2
.Depression was assessed with the Iowa
Short Form (Kohout, Berkman, Evans, & Cornoni-Huntley, 1993) of the Center for
Epidemiological Studies Depression (CES-D) scale (Radloff, 1977). The unifactorial
11-item measure asked participants how frequently they experienced symptoms such
as loss of appetite, changes in sleep patterns, or self-dislike. Coefficient alpha was .90
122 K. Floyd et al.
at T
1
and .89 at T
2
.Stress was measured with the Perceived Stress Scale (PSS)
developed by Cohen, Kamarck, and Mermelstein (1983). Items asked participants
how often, in the past month, they had experienced stress, anger, nervousness,
difficulty coping with irritations, and difficulty dealing with change, among other
symptoms. Coefficient alpha was .90 at T
1
and .90 at T
2
.
Biochemical Analysis
Total serum cholesterol was assessed in mg=dL with the Cholestech LDX, a Clinical Labora-
tory Improvement Amendments (CLIA)-waived in-vitro diagnostic monitor manufac-
turedbyCholestech(Hayward,CA).Forthistest,40ml of capillary blood was
aspirated into a heparinized glass tube and then applied to a sterile test strip that was
assessed by the monitor. During the experiment, the monitor was quantitatively cali-
brated on a daily basis and tested with manufactured controls monthly. Moreover, it
has been extensively validated for cholesterol assessment in published clinical examina-
tions (Gregory, Duh, & Christenson, 1994; Rogers, Misner, Ockene, & Nicolosi, 1993).
Manipulation Checks
A 13-item Likert-type scale was administered on the second and fifth Mondays of the
6-week trial. Participants were sent a link to the questionnaire by e-mail, completed
the questionnaire online, and submitted their answers electronically. When complet-
ing the questionnaire, participants were asked to think specifically about the previous
2 weeks. Each item on the scale was a declarative statement with which participants
were asked to indicate their level of agreement on a scale ranging from 1 (strongly
disagree) to 7 (strongly agree). Embedded within the items was the statement,
‘‘My romantic partner and I have been kissing more than we normally do.’’ This item
provided a direct test of the manipulation. An additional item read, ‘‘My romantic
partner and I have expressed our love for each other verbally more often normal.’’
This item was included both to lessen the uniqueness of the kissing item and to allow
for nonmanipulated changes in verbal affection to be controlled for in the causal ana-
lyses. The remaining items, which were included as potential control variables,
assessed changes in eating patterns (2 items), exercising behaviors (2 items), general
health (2 items), relationship conflict (3 items), and ‘‘filler’’ items related to the cou-
ple’s interaction patterns (2 items). Means and standard deviations (separated by
time and condition) and T
1
–T
2
correlations for all 13 items appear in Table 1.
Results
Manipulation Checks
Scores for each of the 13 manipulation check items were compared by condition. For
the direct manipulation test (‘‘My romantic partner and I have been kissing more
than we normally do’’), those in the experimental group scored significantly higher
than did those in the control group, t(48) ¼5.51, p(one-tailed) <.001, d¼1.59,
indicating success for the manipulation. Concerns that increases in nonverbal
Western Journal of Communication 123
Table 1 Means, Standard Deviations, and T
1
–T
2
Correlations for Items Used in Manipulation Check
Experimental group Control group
Item T
1
M=SD T
2
M=SD T
1
M=SD T
2
M=SD T
1
–T
2
r
I have been exercising less than I normally do. 3.65=2.60 4.31=2.85 4.96=2.60 4.52=2.97 .40
My diet has been healthier than usual. 5.30=1.94 4.15=2.09 5.13=1.98 4.64=2.16 .39
My RM and I have had more conflict than normal. 2.30=1.40 3.38=2.52 2.65=1.87 3.52=2.33 .50
My RM and I have faced a major financial decision. 4.26=2.49 3.12=2.60 4.87=2.63 4.72=3.09 .28
My diet has been less healthy than it usually is. 3.22=1.54 4.27=2.43 3.74=2.14 3.92=1.82 .23
My RM and I have spent more time together than usual. 4.96=1=94 4.27=2.18 5.09=1.90 5.00=2.14 .23
My RM and I have been kissing more than we normally do. 6.30=1.66 3.58=1.84 6.30=1.55 4.80=1.61 .50
I have exercised more often than usual. 5.22=2.19 3.31=2.06 4.74=2.30 4.20=2.55 .25
My RM and I have had difficulty communicating with each other. 2.52=1.41 3.58=2.50 2.39=1.47 3.68=2.39 .54
My health has been better than normal. 4.89=1.69 4.73=1.71 5.22=2.15 4.40=2.12 .19
My RM and I have expressed our love for each other verbally
more often than normal.
5.74=1.86 4.08=1.38 5.48=2.04 5.16=1.89 .18
My RM and I have fought more than we typically do. 2.39=1.41 3.31=2.36 2.48=1.41 3.44=2.20 .34
I haven’t been as healthy as I usually am. 3.26=1.60 3.65=2.43 3.87=2.16 4.20=2.63 .06
Note. Participants were instructed to ‘‘think back over the last 2 weeks’’ and to indicate their level of agreement with each statement. RM ¼romantic partner (this was spelled out
in the measure but is abbreviated here.) Scores ranged from 1 (‘‘strongly disagree’’) to 7 (‘‘strongly agree’’). T
1
–T
2
correlations are averaged across conditions. Significance tests
for correlation coefficients are two-tailed.
p<.05. p<.01.
124
affection would also produce increased verbal affection were assuaged by a nonsigni-
ficant group difference on the item, ‘‘My romantic partner and I have expressed our
love for each other verbally more often than normal.’’
The experimental and control groups differed in their average scores to four other
items measured during the manipulation checks, according to two-tailed tests. One
related to exercise (‘‘I have exercised more often than usual’’), wherein the experi-
mental group reported greater agreement than did the control group, t(48) ¼2.13,
p¼.038, d¼.61. They also differed on all three items related to relational conflict.
For the item, ‘‘My romantic partner and I have had more conflict than we typically
do,’’ the control group exceeded the experimental group, t(48) ¼2.08, p¼.043,
d¼.60. Likewise, for the item, ‘‘My romantic partner and I have had difficulty com-
municating with each other,’’ the control group exceeded the experimental group,
t(48) ¼2.31, p¼.025, d¼.67. Finally, for the item, ‘‘My romantic partner and I
have fought more than we typically do,’’ the control group exceeded the experimental
group, t(48) ¼2.12, p¼.040, d¼.61. Since all 3 conflict items showed a significant
group difference, we aggregated the scores (a¼.91) to create an index of conflict
change, and we used this index, along with the item indicating change in exercise
(described above), as potential covariates.
Descriptive Analyses
T
1
values for all outcome variables (total cholesterol, relationship satisfaction, depres-
sion, and stress) were assessed in sex-by-experimental condition ANOVAs to estab-
lish T
1
equivalency of the experimental conditions. The ANOVAs revealed no main
effects of experimental condition and no sex-by-condition interaction effects for any
of the outcomes (all p’s >.05), indicating equivalency between the experimental and
control conditions. Means and standard deviations for all outcomes, separated by
time and condition, appear in Table 2. There were no significant sex differences in
any of these variables (all p’s >.05).
Table 2 Means and Standard Deviations for Biochemical and Self-Report Outcomes
Variable Condition T
1
MT
1
SD T
2
MT
2
SD
Total cholesterol Experimental 182.56 24.95 176.80 24.46
Control 181.54 26.31 180.69 27.49
Relationship satisfaction Experimental 5.57 1.16 6.18 0.70
Control 6.21 0.78 5.99 0.97
Depression Experimental 2.92 1.36 2.81 0.99
Control 2.30 0.88 2.36 0.84
Stress Experimental 3.57 1.00 2.87 0.81
Control 3.10 1.04 3.15 0.96
Note. Total cholesterol was measured in mg=dL. All self-report outcomes were measured on 7-point scales
wherein higher scores correspond to greater values.
Western Journal of Communication 125
Hypotheses
Analyses and control variables
The hypotheses predicted an increase over time in relationship satisfaction, and
decreases over time in total cholesterol, depression, and stress, in the experimental
condition that are not also observed in the control group. Before testing the hypoth-
eses, we investigated a number of variables as potential covariates, each of which has
been demonstrated to exert independent effects on blood lipids, relationship satisfac-
tion, depression, and=or stress. These potential control variables included: 1) sex; 2)
age; 3) ethnicity; 4) education level; 5) average weekly alcohol consumption; 6) aver-
age weekly caffeine consumption; 7) average number of times per week participant
exercises for at least 30 consecutive minutes; 8) T
1
body mass index; 9) change in
body mass index from T
1
to T
2
; 10) whether or not participant smokes; 11) change
in exercise habits, as measured in manipulation check; and 12) change in conflict
behavior, as measured in manipulation check. All of these variables failed to exert sig-
nificant influence on total cholesterol, relationship satisfaction, depression, and stress
(all p’s >.05). Tests on the remaining outcome variables—since there were no signif-
icant sources of error variance to control—were conducted using ANCOVA with
one-tailed pairwise mean comparisons by condition. Means and standard deviations
for all outcomes, separated by time and condition, appear in Table 2.
Total cholesterol
We examined the effects of the kissing intervention on total cholesterol using
ANCOVA with condition (experimental vs. control) as the fixed factor, T
1
cholesterol
as the covariate, and T
2
cholesterol as the dependent measure. Consistent with the
hypothesis, the ANCOVA produced a significant main effect for condition, F(1, 48) ¼
4.13, p¼.048, partial g
2
¼.08. As expected, the experimental condition experienced a
significant decrease in total cholesterol, t(24) ¼3.34, p¼.001, d¼.64. (Effect size d
reported here and below represents dfor repeated measures, not independent groups.)
Total cholesterol scores in the control group did not differ significantly over time,
t(25) ¼.09, p¼.93.
Relationship satisfaction
An ANCOVA with condition as the fixed factor, T
1
relationship satisfaction as the
covariate, and T
2
relationship satisfaction as the dependent measure produced a
significant main effect for condition, F(1,31) ¼5.53, p¼.025, partial g
2
¼.15. As
hypothesized, the experimental condition experienced a significant increase in
relationship satisfaction, t(15) ¼3.28, p¼.003, d¼.80.
2
Relationship satisfaction
scores in the control group did not differ significantly over time, t(18) ¼1.18, p¼.13.
Depression
An ANCOVA with condition as the fixed factor, T
1
depression as the covariate, and
T
2
depression as the dependent measure produced a nonsignificant main effect for
126 K. Floyd et al.
condition, F(1,31) ¼0.25, p¼.62. Contrary to the prediction, scores for depression
did not differ significantly over time for either the experimental group, t(15) ¼.63,
p¼.27, or the control group, t(18) ¼.59, p¼.28.
Stress
An ANCOVA with condition as the fixed factor, T
1
stress as the covariate, and T
2
stress as the dependent measure produced a significant main effect for condition,
F(1,31) ¼9.47, p¼.004, partial g
2
¼.23. The experimental condition experienced a
significant decrease in stress, t(15) ¼3.13, p¼.004, d¼.78. Stress scores in the
control group did not differ significantly over time, t(18) ¼.28, p¼.39.
Discussion
Theory and previous research suggest that affectionate behavior in personal relation-
ships has stress-ameliorating physiological effects. On this basis, we hypothesized that
increasing affectionate behavior would lead to improvements in physical and mental
health outcomes known to be exacerbated by stress. The present experiment tested
the efficacy of a romantic kissing intervention for improving total cholesterol, depres-
sion, perceived stress, and assessments of relationship satisfaction within cohabiting
marriages and romantic relationships.
As predicted, increased kissing during the 6-week trial preceded statistically signif-
icant decreases in total cholesterol and perceived stress, and a statistically significant
increase in relationship satisfaction, that were not also experienced by those in the
control group. Importantly, these changes cannot be attributed to increased verbal
affection and=or decreased conflict in the experimental group, two additional
communicative changes that one might logically expect to accompany increased non-
verbal affection. It appears, rather, that the parasympathetic effects of kissing in
established, positive-affect relationships account for the observed improvements in
physical, psychological, and relational well-being.
The reduction in total cholesterol is potentially beneficial given that elevated
cholesterol is a primary risk factor for cardiovascular disease (Mader, 2005; Shier
et al., 2004), which is currently the number one cause of mortality for women and
men in the United States (NHLBI, 2008). Previous research by Floyd, Mikkelson,
Hesse, et al. (2007) demonstrated that an affectionate writing exercise was effective
in reducing total cholesterol levels among healthy adults, and the present experiment
replicates this outcome with a nonverbal affection intervention. Considered in
concert, these findings support the contention of affection exchange theory that
affectionate communication has stress-ameliorating physiological effects, one benefit
of which is their ability to improve blood lipid levels.
The experimental condition experienced a significant decrease in perceived stress
and a significant increase in relationship satisfaction not mirrored by the control
group. These results support our theoretic speculation that, whereas low-stress,
high-satisfaction relational environments may increase affectionate behavior among
romantic partners, the obverse causal model is also operative, with increased
Western Journal of Communication 127
affectionate behavior leading to improvements in stress and satisfaction. The
instruction to kiss more frequently has been used as a component of marital therapy
(Brezsnyak & Whisman, 2004), and the present results provide experimental evidence
that such a prescription can enhance relationship satisfaction. The predicted decrease
in depression for the experimental group did not achieve significance, although the
difference was in the hypothesized direction. Previous studies have identified a signif-
icant inverse correlation between affectionate behavior and depression (Floyd, 2002;
Floyd et al., 2005); it may be, however, that depression exerts a stronger causal effect
on affectionate behavior (if there is a causal connection at all) than affectionate
behavior exerts on depression. It may also be the case that assessments of depression
are more complex, and therefore subject to a wider range of potential influences, than
are assessments of satisfaction or stress (which may approximate more general
evaluative judgments). These possibilities await examination in future investigations.
Considered collectively, the present findings suggest that, within the context of
established marital and cohabiting romantic relationships, kissing is a communicative
behavior that effects improvements in some parameters of physical, mental, and rela-
tional well-being. Further, the observed improvements were independent of multiple
potential sources of variance relevant for these outcomes, including body mass index,
exercise and dietary behavior, tobacco use, and caffeine and alcohol consumption.
Along with recent studies by Floyd, Hesse, et al. (2007), Floyd, Mikkelson, Hesse,
et al. (2007), and Floyd, Mikkelson, Tafoya, et al. (2007), this study contributes to
a growing understanding of how interpersonal communication patterns related to
the expression of affection can improve not only self-reported psychological well-
being but also objectively measured markers of physical health, such as blood lipids.
To the extent that increased kissing improves lipid values, it may have utility as an
ancillary nonpharmacological option for treating mild hypercholesterolemia, in con-
cert with traditional interventions, although additional experimental work is war-
ranted before such a course could be recommended.
Limitations and Extensions
Due to the prescreening process and the imposition of multiple inclusion and exclu-
sion criteria, the current sample was probably healthier than a comparably sized sam-
ple drawn at random from the same population would be. This was indicated by
average lipid values that, in both study conditions, were within the range of normal
test values, according to American Heart Association (2007) guidelines. One conse-
quence is that the lipid improvements observed in the experimental group may have
marginal clinical significance for a nonclinical sample such as ours. The extent to
which the observed lipid changes would replicate in a non-laboratory setting is
unknown, but this would be a worthy topic for future field research, given the poten-
tial utility of this intervention as a complementary therapy.
The sample size was small relative to those typically seen in mainstream inter-
personal communication research. It was, however, within the norm both for psycho-
physiological studies (e.g., Kurup & Kurup, 2003; Marazziti & Canale, 2004;
128 K. Floyd et al.
van Niekerk, Huppert, & Herbert, 2001), including those conducted within the field
of interpersonal communication (e.g., Tardy, Thompson, & Allen, 1989). The con-
trolled, longitudinal nature of the current trial, the relative inability of participants
to introduce error variance (at least, in their hematological outcomes) via social
desirability or memory biases, and the emergence of several significant effects all
argue for the adequacy of the sample size.
The items used in the manipulation check measure allowed us to ascertain whether
observed effects might have been attributable to changes in other behaviors besides
kissing, including verbal affection, conflict, and changes in diet or exercise. We were
able to rule out these alternative explanations, but one alternative left unscrutinized is
that increased kissing led to an increase in sexual activity, which could have stress-
alleviating effects on its own and therefore account (at least partially) for the observed
effects. Changes in the frequency of sexual behavior should be ascertained in future
experiments to rule out this alternative explanation.
Future research could extend the present findings in at least four profitable ways.
One would be to test, in a controlled manner, the physiological mechanisms respon-
sible for the observed lipid changes. As we noted in the literature review, the specific
mechanisms via which stress elevates cholesterol are as yet unidentified, but glucocor-
ticoid elevation is one suspect, given that it results directly from arousal of the
hypothalamic-pituitary-adrenal axis, one of the body’s primary physiological
responses to stressors. Future studies should carefully examine this and other poten-
tial mediators of the stress-cholesterol association and as mechanisms through which
reductions in stress can lead to improved lipid values.
Second, the benefits of kissing on physiological, mental, and relational well-being
should be tested within a range of relationship types, including other romantic
relationships (such as lesbian and gay relationships or noncohabiting dyads) and
platonic relationships (such as familial dyads). It is possible that the health benefits
of kissing are limited to romantic kissing, for instance, or that they manifest differ-
ently in romantic and platonic relationships. These and other possibilities await
investigation.
Third, although lipids, depression, perceived stress, and relationship satisfaction
are all theoretically influenced by the experience of stress, they are not the only health
outcomes that are. To the extent that a stress-ameliorating intervention such as affec-
tionate behavior can effect improvements in these outcomes, it may also produce
clinically relevant improvements in other health parameters, such as T lymphocyte
count, natural killer cell cytotoxicity, resting blood pressure, diurnal cortisol varia-
tion, happiness, optimism, and other markers of physical and psychological well-
being. Due to the growing evidence that affectionate behavior has stress-mitigating
effects, future research on these potential outcomes would be warranted.
Finally, given the current intervention’s efficacy with a nonclinical sample, future
research should test its effects in a sample with hypercholesterolemia. To the extent
that it proves effective at improving lipid values for those with chronic high choles-
terol, it may demonstrate clinical as well as statistical significance, further suggesting
its potential utility as a complementary therapy.
Western Journal of Communication 129
Notes
[1] Readers should note that the manipulation check items were scored along a scale of 1 (mean-
ing ‘‘strongly disagree’’) to 7 (meaning ‘‘strongly agree’’), making 4 the theoretic point at
which the participant neither agrees nor disagrees with the statement. Interpreted within this
context, scores on the conflict items (which appear in Table 1) do not indicate increased
conflict – but rather, decreased conflict – for both the experimental and control groups.
Neither group scored even at the midpoint of the scale, on average (i.e., all means were below
4). Thus, it was not the case that people in the control group fought more often than usual –
indeed, their scores indicate that they fought less often than usual – but simply that they did
not decrease their conflict as much as the experimental group did.
[2] Analyses involving the self-report measures of relationship satisfaction, depression, and
stress have fewer degrees of freedom due to missing data at Time 2, when not all participants
completed and returned the questionnaire. We had complete data for these variables at both
time points from 16 experimental participants and 19 control participants, so analyses are
based on these cases only.
References
American Heart Association. (2007). What your cholesterol levels mean. Retrieved June 30, 2007,
from http:==www.americanheart.org=presenter.jhtml?identifier¼183
Bacon, S. L., Ring, C., Lip, G. Y., & Carroll, D. (2004). Increases in lipids and immune cells in
response to exercise and mental stress in patients with suspected coronary artery disease:
Effects of adjustment for shifts in plasma volume. Biological Psychology,65, 237–250.
Banaschak, S., Mo
¨ller, K., & Pfeiffer, H. (1998). Potential DNA mixtures introduced through
kissing. International Journal of Legal Medicine,111, 284–285.
Berscheid, E., Snyder, M., & Omoto, A. M. (1989). The relationship closeness inventory: Assessing
the closeness of interpersonal relationships. Journal of Personality and Social Psychology,57,
792–807.
Brezsnyak, M. & Whisman, M. A. (2004). Sexual desire and relationship functioning: The effects of
marital satisfaction and power. Journal of Sex and Marital Therapy,30, 199–217.
Cane, W. (2005). The art of kissing. (2nd ed.). New York: St. Martin’s Griffin.
Carbary, L. J. (1975). Infectious mononucleosis: The kissing disease. Nursing Care,8, 22–24.
Chrousos, G. & Gold, P. (1992). The concepts of stress and stress system disorders. Journal of the
American Medical Association,267, 1244–1252.
Cohen, S., Kamarck, T., & Mermelstein, R. (1983). A global measure of perceived stress. Journal of
Health and Social Behavior,24, 385–396.
Comfort, A. (1974). The likelihood of human pheromones. In M. C. Birch (Ed.), Pheromones
(3rd ed., pp. 386–394). North-Holland: Amsterdam.
Cowan, F. M., Copas, A., Johnson, A. M., Ashley, R., Corey, L., & Mindel, A. (2002). Herpes
simplex virus type 1 infection: A sexually transmitted infection of adolescence? Sexually
Transmitted Infections,78, 346–348.
De Chateau, P. & Wiberg, B. (1977). Long-term effect on mother-infant behavior of extra contact
during the first hour post partum: II. A follow-up at three months. Acta Paediatrica
Scandinaica Supplement,66, 145–151.
Eibl-Eibesfeldt, I. (1970). Love and hate: On the natural history of behavior patterns. New York:
Methuen.
Floyd, K. (2002). Human affection exchange: V. Attributes of the highly affectionate. Communica-
tion Quarterly,50, 135–152.
Floyd, K. (2006a). Communicating affection: Interpersonal behavior and social context. Cambridge,
England: Cambridge University Press.
130 K. Floyd et al.
Floyd, K. (2006b). Human affection exchange: XII. Affectionate communication is associated with
diurnal variation in salivary free cortisol. Western Journal of Communication,70, 47–63.
Floyd, K., Hess, J. A., Miczo, L. A., Halone, K. K., Mikkelson, A. C., & Tusing, K. J. (2005). Human
affection exchange: VIII: Further evidence of the benefits of expressed affection. Communica-
tion Quarterly,53, 285–303.
Floyd, K., Hesse, C., & Haynes, M. T. (2007). Human affection exchange: XV. Metabolic and
cardiovascular correlates of trait expressed affection. Communication Quarterly,55, 79–94.
Floyd, K., Mikkelson, A. C., Hesse, C., & Pauley, P. M. (2007). Affectionate writing reduces total choles-
terol: Two randomized, controlled trials. Human Communication Research,33, 119–142.
Floyd, K., Mikkelson, A. C., Tafoya, M. A., Farinelli, L., La Valley, A. G., Judd, J., et al. (2007).
Human affection exchange: XIII. Affectionate communication accelerates neuroendocrine
stress recovery. Health Communication,22, 123–132.
Floyd, K. & Morman, M. T. (1998). The measurement of affectionate communication. Communica-
tion Quarterly,46, 144–162.
Giese, D. J., Sephton, S. E., Abercrombie, H. C., Duran, R. E. F., & Spiegel, D. (2004). Repression
and high anxiety are associated with aberrant diurnal cortisol rhythms in women with
metastatic breast cancer. Health Psychology,23, 645–650.
Greenbaum, P. E. & Rosenfeld, H. M. (1980). Varieties of touching in greetings: Sequential
structure and sex-related differences. Journal of Nonverbal Behavior,5, 13–25.
Gregory, L. C., Duh, S. H., & Christenson, R. H. (1994). Eight compact analysis systems evaluated
for measuring total cholesterol. Clinical Chemistry,40, 579–585.
Grewen, K. M., Girdler, S. S., Amico, J., & Light, K. C. (2005). Effects of partner support on resting
oxytocin, cortisol, norepinephrine, and blood pressure before and after warm partner
contact. Psychosomatic Medicine,67, 531–538.
Guerrero, L. K. & Floyd, K. (2006). Nonverbal communication in close relationships. Mahwah, NJ:
Lawrence Erlbaum Associates.
Heim, C., Ehlert, U., & Hellhammer, D. H. (2000). The potential role of hypocortisolism in the
pathophysiology of stress-related bodily disorders. Psychoneuroendocrinology,25, 1–35.
Hendrick, S. S. (1988). A generic measure of relational satisfaction. Journal of Marriage and the
Family,50, 93–98.
Huston, T. L., Caughlin, J. P., Houts, R. M., Smith, S. E., & George, L. J. (2001). The connubial
crucible: Newlywed years as predictors of marital delight, distress, and divorce. Journal of
Personality and Social Psychology,80, 237–252.
Huttunen, J. K., Ehnolm, C., Kekki, M., & Nikkila, E. A. (1976). Post-heparin plasma lipoprotein
Lipase and hepatic lipase in normal subjects and in patients with hypertriglyceridaemia:
Correlations to sex, age, & various parameters of triglyceride metabolism. Clinical Science
and Molecular Medicine,50, 249–260.
Jansen, H. & Hulsmann, W. C. (1985). Enzymology and physiological role hepatic lipase. Biochemical
Society Transactions,3, 24–26.
Kimata, H. (2003). Kissing reduces allergic skin wheal responses and plasma neurotrophin levels.
Physiology and Behavior,80, 395–398.
Kimata, H. (2006). Kissing selectively decreases allergen-specific IgE production in atopic patients.
Journal of Psychosomatic Research,60, 545–547.
Kirschbaum, C. & Hellhammer, D. H. (1989). Salivary cortisol in psychobiological research: An
overview. Neuropsychology,22, 150–169.
Kohout, F. J., Berkman, L. F., Evans, D. A., & Cornoni-Huntley, J. (1993). Two shorter forms of the
CES-D depression symptoms index. Journal of Aging and Health,5, 179–193.
Kurup, R. K. & Kurup, P. A. (2003). Hypothalamic digoxin, hemispheric dominance, and
neurobiology of love and affection. International Journal of Neuroscience,113, 721–729.
Leonard, T. L. (1990). Male clients of female street prostitutes: Unseen partners in sexual disease
transmission. Medical Anthropology Quarterly,4, 41–55.
Western Journal of Communication 131
Liccardi, G., Gilder, J., D’Amato, M., & D’Amato, G. (2002). Drug allergy transmitted by passionate
kissing. Lancet,359, 1700.
Light, K. C., Grewen, K. M., & Amico, J. (2005). More frequent partner hugs and higher oxytocin
levels are linked to lower blood pressure and heart rate in premenopausal women. Biological
Psychology,69, 5–21.
Lui, W. & Tang, H. (2004). Preliminary study on electrotactile sensitivity of the lips. Engineering in
Medicine and Biology Society,2, 4811–4814.
Mader, S. S. (2005). Understanding human anatomy and physiology (5th ed.). New York:
McGraw-Hill.
Maged, A. (2006, November 20). With flu season in swing, an innocent kiss may be harmful. The
Daily Star=Egypt. Retrieved June 22, 2007, from http:==www.dailystaregypt.com=
article.aspx?ArticleID¼4047
Mancuso, G. & Berdondini, R. M. (2006). Kiss-induced allergy to amoxicillin. Contact Dermatitis,
54, 226.
Maloney, J. M., Chapman, M. D., & Sicherer, S. H. (2006). Peanut allergen exposure through saliva:
Assessment and interventions to reduce exposure. Journal of Allergy and Clinical Immunol-
ogy,118, 719–724.
Marazziti, D. & Canale, D. (2004). Hormonal changes when falling in love. Psychoneuroendocrinol-
ogy,29, 931–936.
McCall, R. E. & Tankersley, C. M. (2003). Phlebotomy essentials (3rd ed.). Philadelphia: Lippincott,
Williams & Wilkins.
McCann, B. S., Magee, M. S., Broyles, F. C., Vaughan, M., Albers, J. J., & Knopp, R. H. (1995). Acute
psychological stress and epinephrine infusion in normolipidemic and hyperlipidemic men:
Effects of plasma lipid and apoprotein concentrations. Psychosomatic Medicine,57, 165–176.
Miller, W. C., Gorski, J., Oscai, L. B., & Palmer, W. K. (1989). Epinephrine activation of heparin
non-releasable lipoprotein lipase in 3 skeletal muscle fiber types of the rat. Biochemical and
Biophysical Research Communications,164, 615–619.
Montagna, W. & Parakkal, P. F. (1974). The structure and function of skin. New York: Academic
Press.
Muldoon, M. F., Herbert, T. B., Patterson, S. M., Kameneva, M., Raible, R., & Manuck, S. B. (1995).
Effects of acute psychological stress on serum lipid levels, hemoconcentration, and blood
viscosity. Archives of Internal Medicine,155, 615–620.
National Heart. Lung. and Blood Institute. (NHLBI). (2008). High blood cholesterol: What you
need to know. Retrieved January 5, 2008, from http://www.nhlbi.nih.gov=health=
public=heart=chol=wyntk.htm
Nicholson, B. (1984). Does kissing aid human bonding by semiochemical addiction? British Journal
of Dermatology,111, 623–627.
Owen, W. F. (1987). The verbal expression of love by women and men as a critical communication
event in personal relationships. Women’s Studies in Communication,10, 15–24.
Penn, M. (2003). Ritual kissing, heresy and the emergence of early Christian orthodoxy. Journal of
Ecclesiastical History,54, 625–640.
Piazza, M., Chirianni, A., Picciotto, L., Guadagnino, V., Orlando, R., & Cataldo, P. T. (1989). Pas-
sionate kissing and microlesions of the oral mucosa: Possible role in AIDS transmission.
Journal of the American Medical Association,261(2), 244–245.
Radloff, L. S. (1977). The CES-D scale: A self-report depression scale for research in the general
population. Applied Psychological Measurement,1, 385–401.
Rogers, E. J., Misner, L., Ockene, I. S., & Nicolosi, R. J. (1993). Evaluation of seven Cholestech LDX
analysers for total cholesterol determinations. Clinical Chemistry,39, 860–864.
Rubinsky, H. J., Hoon, P. W., Eckerman, D. A., & Amberson, J. I. (1985). Groin skin temperature:
Testing the validity of a relatively unobtrusive physiological measure of psychosexual arousal.
Psychophysiology,22, 488–492.
132 K. Floyd et al.
Schoch-Spana, M. (2000). Implications of pandemic influenza for bioterrorism response. Clinical
Infectious Diseases,31, 1409–1413.
Seeley, T., Talbot, T., Abramson, P. R., Perry, L. B., Rothblatt, A. B., & Seeley, D. M. (1980).
Thermographic measurement of sexual arousal: A methodological note. Archives of Sexual
Behavior,9, 77–86.
Shier, D., Butler, J., & Lewis, R. (2004). Hole’s human anatomy and physiology (10th ed.). New York:
McGraw-Hill.
Stenn, K. S. & Bhawan, J. (1990). The normal histology of the skin. In E. R. Farmer & A. F. Hood
(Eds.), Pathology of the skin (pp. 122–164). Norwalk, CT: Appleton and Lange.
Steptoe, A. & Brydon, L. (2005). Associations between acute lipid stress responses and fasting lipid
levels 3 years later. Health Psychology,24, 601–607.
Stoney, C. M., Niaura, R., Bausserman, L., & Metacin, M. (1999). Lipid reactivity to stress: I. Com-
parison of chronic and acute stress response. Health Psychology,18, 241–250.
Tardy, C. H., Thompson, W. R., & Allen, M. T. (1989). Cardiovascular responses during speech:
Does social support mediate the effects of talking on blood pressure? Journal of Language
and Social Psychology,8, 271–285.
Tully, J., Viner, R. M., Coen, P. G., Stuart, J. M., Zambon, M., Peckham, C., et al. (2006, February
10). Risk and protective factors for meningococcal disease in adolescents: Matched cohort
study. British Medical Journal Online. Retrieved June 22, 2007, from http:==www.bmj.com=
cgi=rapidpdf=bmj.38725.728472.BEv1
Turner, R. A., Altemus, M., Enos, T., Cooper, B., & McGuinness, T. (1999). Preliminary research on
plasma oxytocin in normal cycling women: Investigating emotion and interpersonal distress.
Psychiatry,62, 97–113.
Uvna
¨s-Moberg, K. (1997). Physiological and endocrine effects of social contact. Annals of the
New York Academy of Sciences,807, 146–163.
van Niekerk, J. K., Huppert, F. A., & Herbert, J. (2001). Salivary cortisol and DHEA: Association
with measures of cognition and well-being in normal older men, and effects of three months
of DHEA supplementation. Psychoneuroendocrinology,26, 591–612.
Waring, E. M. (1984). The measurement of marital intimacy. Journal of Marital and Family
Therapy,10, 185–192.
Wedekind, C., Seebeck, T., Bettens, F., & Paepke, A. J. (1995). MHC-dependent mate preferences in
humans. Proceedings: Biological Sciences,260, 245–249.
Willett, E. (2001). The science of kissing. Retrieved May 20, 2007, from http:==www.edwardwillett.
com=Columns=kissing.htm
Western Journal of Communication 133
