Content uploaded by Charles J Wysocki
Author content
All content in this area was uploaded by Charles J Wysocki on Dec 10, 2015
Content may be subject to copyright.
Research Article
Preference for Human Body
Odors Is Influenced by Gender
and Sexual Orientation
Yolanda Martins,
1,2
George Preti,
1,3
Christina R. Crabtree,
4
Tamar Runyan,
1
Aldona A. Vainius,
1
and
Charles J. Wysocki
1,5
1
Monell Chemical Senses Center, Philadelphia, Pennsylvania;
2
School of Psychology, Flinders University of South
Australia, Adelaide, Australia;
3
Department of Dermatology, University of Pennsylvania;
4
Hobart and William Smith
Colleges; and
5
Department of Animal Biology, School of Veterinary Medicine, University of Pennsylvania
ABSTRACT—Human body odor may contribute to selection
of partners. If so, sexual orientation may influence pref-
erence for and perhaps production of human body odors.
In a test of these hypotheses, heterosexual and homosexual
males and females made two-alternative forced-choice
preference judgments for body odors obtained from other
heterosexual and homosexual males and females. Subjects
chose between odors from (a) heterosexual males and gay
males, (b) heterosexual males and heterosexual females,
(c) heterosexual females and lesbians, and (d) gay males
and lesbians. Results indicate that differences in body odor
are detected and responded to on the basis of, in part, an
individual’s gen der and sexual orientation. Possible mech-
anisms underlying these findings are discussed.
Pheromones are chemical signals that influence the behavior or
physiology of other organisms (Karlson & Luscher, 1959). Evi-
dence for such chemical communication in humans has been
reported (Preti & Wysocki, 1999; Preti, Wysocki, Barnhart,
Sondheimer, & Leyden, 2003). Considerable information shows
that humans produce individually unique body odors (odor
prints) that are determined in large measure by the set of linked
genes that regulate the immune system, are perceptible to oth-
ers, and may act as pheromones. In part, odor prints allow in-
dividuals to identify subgroups within the population, which
suggests odor prints have a role in social recognition (Porter,
Balogh, Cernoch, & Franchi, 1986; Porter & Winberg, 1999;
Russell, 1976; Singh, 2001). Supporting evidence for this pos-
sibility comes from earlier studies demonstrating that humans
can easily discriminate among body odors from different people.
For example, individuals are capable of distingui shing kin and
nonkin, or their own odors and those of other people (Porter et
al., 1986; Russell, 1976). Even newborn infants are more likely
to orient to odors donated from the breast of their own mother
rather than an unfamiliar mother; thus, it appears that the ability
to detect differences in human body odors begins early in life
(Porter & Winberg, 1999).
In many mammals, pheromones modulate, and sometimes
determine, reproductive behaviors (Vandenbergh, 1989; Wyatt,
2003). For example, odors of the opposite sex are highly at-
tractive, especially for males when females are in heat (Nyby,
1982; Rasmussen, Lee, Roelofs, Zhang, & Daves, 1996). If
humans use the odors of others, at least in part, to select a mate
(Jacob, McClintock, Zelano, & Ober, 2002; Ober, Weitkamp, &
Cox, 1999; Ober et al., 1997; Weitkamp & Ober, 1999), then
sexual orientation may influence preferences for human odors or
production of these odors.
Although 2 to 10% of the population is lesbian or gay male
(Binson et al., 1995; Kinsey, Pomeroy, & Martin, 1948; Lau-
mann, Gagnon, Michael, & Michaels, 1994; Sell, Wells, &
Wypij, 1995), prior investigations of human body odors have
been carried out without regard for the sexual orientation of
participants. We sought to determine whether there is a differ-
ence in preference for axillary odor from heterosexuals, lesbi-
ans, and gay men and whether preference for these odors is
influenced by an odor evaluator’s gender and sexual orientation.
In this study, heterosexuals, lesbians, and gay men indicated
their preference for the odors of underarm sweat collected from
other heterosexuals, lesbians, and gay males.
Address correspondence to Charles J. Wysocki, Monell Chemical
Senses Center, 3500 Market St., Philadelphia, PA 19104, e-mail:
wysocki@monell.org, or to Yolanda Martins, e-mail: yolanda.martins@
flinders. edu.au.
PSYCHOLOGICAL SCIENCE
694 Volume 16—Number 9Copyright r 2005 American Psychological Society
METHOD
Odor Donors
Odor donors were 24 individuals (age range 5 18–35), 6 from
each of the following four categories: heterosexual males, gay
males, heterosexual females, and lesbians. Participants were
recruited from the University of Pennsylvania and the greater
Philadelphia community. Lesbians and gay males were recruited
from gay and lesbian groups in the community or were friends of
participants who were recruited from these groups.
1
Potential donors rated their current sexual behavior (past 12
months) on the 7-point Kinsey Scale (Kinsey et al., 1948),
ranging from 0, exclusive ly heterosexual with no homosexuality,
to 6, exclusively homosexual with no heterosexuality. Only those
individuals describing their current sexual behavior as exclu-
sively heterosexual, lesbian, or gay were invited to participate
in the study proper. Participants received $75 for completing
the odor-donation phase.
Donors underwent a 9-day ( 2 days) wash-out phase during
which they used odorless soaps and shampoo, did not shave their
armpits, and eliminated items such as garlic, cumin, and curry
from their diets. After this phase, donors wore cleaned, cotton
gauze pads in the armpits throughout the day for 3 days,
changing to a new pad each day (for details about this procedure,
see Preti, Cutler, Huggins, Garcia, & Lawley, 1986), while
taking part in their normal daily activities. Although the envi-
ronment that individuals were in (e.g., work, school) while col-
lections took place was not controlled, the specific environment
from which the odors were collected was controlled (i.e., an
unshaven armpit that had not been washed since the previous
evening). Any differences in the odor collections due to differ-
ences in lifestyle activities were likely negligible and randomly
distributed across all groups of odor donors. Pads were kept in
participants’ freezers in cleaned, labeled glass jars (provided by
us) until their return to the laboratory. The pads were then frozen
at 80 1C, to maintain the integrity of the chemosignal. After all
pads were returned to the laboratory, each pad was cut into six
pieces. Each test bottle was created by taking two pieces from
each of three donors within a group and combining them in an
odorless, plastic squeeze bottle with a flip-top lid. Hence, we
created unique odors within each odor class, so that no indi-
vidual odor would be recognized or exert undue influence on our
results. The selection of donor pads to be used in each squeeze
bottle was randomized, with the constraint that all donors be
represented in the final set of test bottles at approximately the
same frequency. Test bottles were used for a maximum of two
testing sessions.
Odor Evaluation
Odor Evaluators
Eighty-two individuals were odor evaluators (age range 5 18–
35). Data from 2 participants were excluded because of com-
puter difficulties during testing, leaving 20 participants in each
group. Mean ages (standard deviations in parentheses) were 24
( 5) for heterosexual males, 25 ( 5) for gay males, 23 ( 4)
for heterosexual females, and 25 ( 5) for lesbians. Recruit-
ment and inclusion criteria were identical to those described for
the odor donors. Participants received $15/hr for completion of
this phase.
Odor Judgments
Evaluators made forced-choice preference judgments for odors
from each of the following pairings: (a) heteros exual male versus
gay male, (b) heterosexual male versus heterosexual female, (c)
heterosexual female versus lesbian, and (d) gay male versus
lesbian. Odor pairs were presented at 1-min intertrial intervals,
11 times each, counterbalanced across participants, for a total of
44 trials. Thus, for each testing session, there were two bottles
from each odor class. The recombined unique odors in the two
bottles from each odor class were identical within a given testing
session. Order of presentation of a partic ular odor within a given
pair was counterbalanced across trials for each participant. In
addition, strength of preference (on a 10-point scale) was col-
lected for the preferred odor in each trial.
Following the preference trials, participants rated the inten-
sity (on a mod ified labeled-magnitude scale—Green, Shaffer, &
Gilmore, 1993) and then the pleasantness (on a 23-point bipolar
scale) of the odors, two times for each of the odors presented in
the forced-choice trials, for a total of four ratings per odor class.
Order of presentation of odors was counterbalanced across
participants.
Procedure
Participants were tested individually and were informed that the
purpose of the experiment was to assess their preference for
body odors. They were told that odors would be presented in
pairs, in unlabeled, plastic squeeze bottles with flip-top lids, and
that their task would be to indicate which odor-containing bottle
they preferred. In addition, they were told that they would be
presented with 44 of these pairs. They were then instructed in
the standardized method for sniffing the odor from each test
bottle: shake the bottle vigorously, flip the lid, take a ‘‘hearty’’
sniff while holding the bottle just below the nostrils, and close
the lid tightly after one sniff. Participants completed 22 forced-
choice preference trials, received a 10-min break, and then
1
We did not assess the AIDS-HIV status of our odor donors, and it is possible
that individuals who have AIDS-HIV smell different from those who do not have
this disease, in part because of their compromised immune systems. Such a
difference could have affected the results we describe in this report. We have
attempted to determine what the AIDS-HIV rates are in the populations from
which we sampled odor donors, but have learned that accurate data do not exist.
It should be noted, however, that the odor stimuli used over the course of the
study were prepared in a manner designed to minimize the influence of any single
donor (see the Method section for details). Therefore, we believe it unlikely that
the inclusion of an HIV-positive individual among the donors would have pro-
duced a systematic bias responsible for the results.
Volume 16—Number 9 695
Y. Martins et al.
completed the remaining 22 trials. After an additional 7-min
break, they returned to rate the intensity and pleasantness of
each odor.
2
RESULTS
Forced-Choice Preferences
For each odor pair tested, we calculated the total number of
times (over all 11 trials per pair) that a particular odor was
chosen. These data were then averaged for each gender and
sexual-orientation group. If there were no detectable differences
in the body odors across the donor groups, one would expect the
evaluators to have determined their preferences by chance, with
participants choosing one odor in half of the forced-choice
preference trials and the competing odor in the other half of the
trials. Our data refute this outcome: Participants clearly ex-
hibited preferences for specific odors at greater than chance
levels (se e Fig. 1).
It should be noted that all analyses were first carried out on all
lesbian participants, and then separate analyses were performed
for lesbians from ages 18 to 25 and those over the age of 25. We
did this becau se prior research on lesbians has demonstrated
that the development of a lesbian orientation is qualitatively
different from the development of sexual orientation in other
groups (i.e., heterosexual males and females and gay males).
More specifically, lesbians begin questioning their sexual ori-
entation later than othe r individuals and are more likely to begin
questioning their individual identity because of ideological
beliefs or exposure to lesbian sexua l orientation than because of
sexual or emotional attraction to same-sex others. Essentially, a
lesbian orientation is typically established later than the sexual
orientation of other groups (Diamond, 1998; Henderson, 1984;
Kitzinger & Wilkinson, 1995; Rust, 1993). The only resu lt that
changed when the data were analyzed separately for the two age
groups was the forced-choice preference between odors from gay
males and odors from lesbian females.
Heterosexual males, heterosexual females, and lesbians
preferred odors from heterosexual males over odors from gay
males (Fig. 1); gay males preferred odors from other gay males
(all comparisons significant at p < .01). Heterosexual males,
heterosexual females, and lesbians over the age of 25 (but not
those ages 18–25) preferred odors from lesbians over odors from
gay males (Fig. 1); gay males prefer red the odors of other gay
males (all comparisons significant at p < .01). In trials com-
paring odors from heterosexual females and lesbians, all but
heterosexual males significantly preferred the odors of hetero-
sexual females (Fig. 1). Finally, gay males preferred odors from
heterosexual females over those from heteros exual males (Fig.
1); the other groups of evaluators showed no significant prefer-
ence for either one of these odors.
Fig. 1. Results for the two-alternative forced-choice preference judg-
ments. The key in the upper half of the figure explains how to interpret the
results for the four groups of odor evaluators, seen in the lower half of the
figure. Results are represented by circles in a square array, with the po-
sition of each circle indicating the odor-donor group and the lines between
neighboring circles (dotted vs. solid) indicating whether the comparison
between those two odors showed that evaluators had a significant prefer-
ence for one over the other. There are five possible circle sizes, though only
four are used. If the odor of a donor group was significantly nonpreferred
in comparisons with the odors of the two groups at the neighboring cor-
ners, it is represented by the smallest circle (score of 2). If the odor of a
donor group was significantly preferred over the odors of the two groups at
the neighboring corners, it is represented by the largest circle (score of
12). Intermediate-size circles represent scores of 11 (significantly pre-
ferred over one of the comparison odors, but not both) and 0 (either sig-
nificantly different from both comparison odors, but in opposite
directions—indicated by an asterisk—or not significantly different from
either comparison odor). The plus sign indicates that analyses were per-
formed using data only from lesbians over the age of 25 (see the text for the
rationale).
2
Because of a computer malfunction, these ratings were not collected from the
first 15 participants.
696 Volume 16—Number 9
Preference for Human Body Odors
Overall, Figure 1 illustrates some striking global tendencies.
First, it is clear that gay males, as odor evaluators, are strikingly
different from heterosexual males and females, as well as les-
bians: Gay males preferred the odors of gay males and hetero-
sexual females in each of the forced-choice preference tests in
which these odors were present. In contrast, for heterosexual
males, heterosexual females, and lesbians, odors from gay males
were the most nonpreferred odor presented. Of further interest is
the equivalence in the pattern of responses for heterosexual
females and lesbians. Both grou ps preferred odors from het-
erosexuals of either gender over odors from gay males and les-
bians, but when forced to choose between the latter, they pre-
ferred odors from lesbians over those of gay males. Heterosexual
males showed a similar, though not equivalent, pattern of re-
sponding: They exhibited no preferences in choices between the
odors from other heterosexual males versus those from hetero-
sexual females, or between the odors from the two female groups.
They did, however, clearly prefer the odor from lesbians over the
odor from gay male donors.
Table 1 presents the number of participants who chose a
specific odor within each odor pair on a significant proportion of
the trials (i.e., at least 9 of the 11 trials). These data provide
TABLE 1
Mean Strength-of-Preference (SoP) Ratings (With Standard Deviations) and Number of
Participants Choosing a Specific Odor a Significant Number of Times
Evaluator group
Choice and statistic
Heterosexual
males
Gay
males
Heterosexual
females
Lesbian
females
Heterosexual male vs. gay male
Chose heterosexual male
Mean SoP 4.63 2.17 6.11 5.35
SD 2.24 1.47 2.95 2.48
Number of evaluators 15 3 14 11
Chose gay male
Mean SoP 4.03 5.63 3.98 3.54
SD 3.17 1.77 0.82 4.77
Number of evaluators 2 8 2 6
Heterosexual male vs. heterosexual female
Chose heterosexual male
Mean SoP 1.57 0.42 3.13 2.51
SD 1.15 — 2.53 1.74
Number of evaluators 6 1 4 6
Chose heterosexual female
Mean SoP 4.85 4.32 0.96 6.04
SD 2.15 3.29 5.61 2.10
Number of evaluators 2 8 3 7
Lesbian female vs. heterosexual female
Chose lesbian female
Mean SoP 1.16 6.17 2.55 2.86
SD 0.64 4.26 — —
Number of evaluators 2 2 1 1
Chose heterosexual female
Mean SoP 2.47 5.20 3.48 4.27
SD 3.70 2.69 3.76 2.33
Number of evaluators 7 10 9 9
Gay male vs. lesbian female
Chose gay male
Mean SoP — 4.90 4.04 4.07
SD — 1.02 — 1.42
Number of evaluators 0 8 1 4
Chose lesbian female
Mean SoP 4.62 0.59 5.72 6.39
SD 2.16 7.64 2.38 1.75
Number of evaluators 12 3 15 8
Volume 16—Number 9 697
Y. Martins et al.
evidence that many participa nts reliably chose the same odor
within each pair when tested over repe ated trials. In many in-
stances, the frequency of participants with a significant choice
bias is greater than would be expected by chance. With an alpha
level of .05 and a sample size of 20 in each group, one would
expect that, by chance alone, only 1 participant per group would
have a significant choice bias for each specific pair of odors.
Moreover, examination of the strength-of-preference data (Table
1) shows that, often, when many participants significantly chose
a given odor, as a group they tended to have a high strength-of-
preference rating for that odor (these data could not be subjected
to infere ntial statistical tests because some cells have small ns).
Frequently, differences in strength-of-preference ratings were
related to participa nts’ gender and sexual orientation and to the
odor that was chosen (see Table 1). For example, in the com-
parison of odors from gay males and odors from heterosexual
males, the preference of gay males who repeatedly chose odors
from other gay males was stronger than the preference of gay
males who repeatedly chose odors from heterosexua l males.
Similarly, in the comparison of odors from heterosexual males
and odors from heterosexual females, the preference of hetero-
sexual females who chose odors from heterosexual males was
moderately stronger than the preference of heterosexual females
who chose odors from heterosexual females. Furthermore, het-
erosexual males, heterosexual females, and lesbians who reli-
ably chose gay male odors exhibited a weaker strength of
preference for these odors than did heterosexual males, hetero-
sexual females, and lesbians who reliably chose the comparison
odors (heterosexual male or lesbian female).
To explore the relationship between mean strength of prefer-
ence for a given odor and the percentage of times that odor was
chosen, we calculated the correlations between these two vari-
ables for each forced-choice comparison pair as a function of
participants’ gender and sexual orientation. To do this, we
converted mea n strength of prefer ence to a bipolar measure for
each pair, by multiplying the strength-of-preference rating for
the outcome of certain subsets of choices by 1. For example,
the strength-of-preference ratings for heterosexual male odors in
the comparisons of heterosexual and gay males and of hetero-
sexual males and females were multiplied by 1. The correla-
tions were all highly significant (all ps < .001) and ranged from
.76 to .95, indicating that as the strength of preference for a
given odor increased, the percentage of times that odor was
chosen in the forced-choice preference test also increased.
Overall, these data suggest that lesbians and gay males may
produce an array of axillary odorants that distinguish them from
heterosexuals. Furthermore, gay males may perceive these charac-
teristic odorants differently than their heterosexual counterparts do.
Perceived Intensity
To rule out the possibility that individuals’ perceptions of the
intensity of the different odors accounted for these results, we
conducted a 4 (participant’s group membership) 4 (odor class)
split-plot analysis of covariance (ANCOVA) on individuals’
ratings of intensity of the odors (see footnote 2). Participant’s age
was the covariate. This analysis yielded no significant effects,
indicating that participants’ perceptions of the intensity of the
different odors did not differ across the groups of participants,
F(3, 60) 5 0.543, n.s., or across the odor classes, F(2, 107) 5
2.23, n.s., either alone or in interaction, F(5, 107) 5 1.81, n.s.
Perceived Pleasantness
We also performed a 4 4 split-plot ANCOVA on individuals’
pleasantness ratings of the odors.
3
This analysis revealed that
participants’ perceptions of the pleasantness of the odors varied
depending on the interaction between participants’ group mem-
bership and the odor class, F(6, 126) 5 2.24, p < .05, Z
p
2
5 .10.
No other significant effects were found. These results suggest
that odor choice is driven primarily by perceptions of perceived
odor pleasantness or unpleasantness and that the perceived
pleasantness is determined, in part, by the perceiver’s gender
and sexual orientation (Fig. 2).
DISCUSSION
In the present work, body odors obtained from heterosexual and
homosexual males and females generated patterns of prefer-
ences that differed across groups of individuals having different
sexual orientations. These differences were not accounted for by
individuals’ perceptions of the intensity of the odors. Recently,
Savic, Berglund, and Lindstro
¨
m (2005) reported similar patterns
of brain activation in gay males and heterosexual females, which
differed from the pattern seen in heterosexual males, while they
smelled high concentrations of the odorants estra-1,3,5(10), 16-
tetraen-3-ol and 4-16-androstadien-3-one (AND); lesbians were
not examined. AND is found in the underarm sweat of both males
and females. Together, these studies suggest that there are
qualitative differenc es in the perception of axillary odors that
are correlated with gender and sexual orientation. Differences in
the production or perception of body odor may result from en-
vironmental or biological factors (e.g., familiarity with the odors,
lifestyle factors such as frequency of exercising, differences in
axillary flora, or differences in genetic alleles).
There is clear evidence to support the assertion that odor
prints allow individuals to identify subgroups or classes within
the population, which suggests that odor prints have a role in
social recognition (Porter et al., 1986). Research in chemical
communication has determined that individuals’ odor prints are
primarily established by the major histocompatibility complex
(MHC; also called human leukocyte antigen, HLA, in people), a
3
For these analyses, F tests involving the odor-class variable were calculated
using the Greenhouse-Geisser correction because Mauchly’s test of sphericity
was significant for this variable, W(5) 5 .32 for intensity and .51 for pleasant-
ness, ps < .01.
698 Volume 16—Number 9
Preference for Human Body Odors
linked set of genes that regulates the immune system (for re-
views, see Singh, 2001; Yamazaki, Singer, Curran, & Beau-
champ, 1999; and Yamazaki, Singer, & Beauchamp, 1998–
1999). Historically, studies have focused on the relationships
among an individual’s odor preference, his or her HLA type,
the HLA types of the odor donors, and, in at least one report,
the HLA types of the odor evaluator’s parents (Jacob et al.,
2002; Ober et al., 1997, 1999; Weitkamp & Ober, 1999). These
studies have shown that individuals tend to prefer odors from
other individuals who have some HLA alleles identical to their
own, but avoid odors from individuals with only one or no allele
matches, as well as individuals with identical HLA alleles.
Furthermore, these experiments provide initial evidence that
the MHC may play a role in mate choice among humans (Ober
et al., 1997, 1999), as it does in other mammals (Yamazaki et al.,
1999).
Interestingly, research examining the biological mechanisms
underlying human sexual orientation has implied that there may
be a relationship between the MHC and sexual orientation. The
most pervasive (and influential) statemen t regarding sexual
orientation is the neurohormonal hypothesis (Ellis & Ames,
1987), according to which homosexuality results from exposure
to atypical androgen levels in utero. Although evidence for this
hypothesis has been inconsistent, some work has illustrated that
heterosexual and homosexual men and wome n differ on a variety
of characteristics known or believed to affect androgens (e.g.,
neuroanatomy and anthropometrics ; for a review, see Mustanski,
Chivers, & Bailey, 2002). Recently, a genome scan, focusing on
gay males, suggested genetic regions that contained candidate
genes involved in the regulation of hormones thought to be in-
volved in sexual orientation (Mustanski et al., 2005). Further
support for the neurohormonal hypothes is comes from Gesch-
wind and Galaburda’s (1985) criticized theory of cerebral lat-
eralization and Ivanyi’s work (1978) on the MH C of the mouse.
These researchers have suggested that levels of sensitivity to
testosterone are influenced by genes in the MHC. Indeed, Iv-
anyi’s work has illustrated that loci found in the MHC cont rol
(either directly or indirectly) myriad testosterone-related
mechanisms, including serum levels of and responsiveness to
testosterone and levels of testosterone-binding globulin. Thus,
differences in MHC alleles that differentially influence sexual
orientation may be related to differences in MHC-regulated odor
production.
Studies have shown that among men, sexual orientation cor-
relates with an individual’s number of older brothers. This line
of research has suggested a link between homosexuality and
Fig. 2. Mean pleasantness ratings of each odor class as a function of odor evaluators’ group
membership.
Volume 16—Number 9 699
Y. Martins et al.
the minor histocompatibility antigens, at least among men.
Blanchard and Bogaert (1996), the primary proponents of this
work, have suggested that the correlation may be explained by
what is aptly termed the maternal immunization hypothesis
(Blanchard, 2001). According to this hypothesis, mothers are
progressively immunized against the Y-linked minor histo-
compatibility antigens (i.e., H-Y antigens) expressed by male
(but not female) fetuses. In turn, these antibodies prevent the
fetal brain from developing in a masculine fashion, instead
causing it to be feminized. This hypothesis provides a parsi-
monious explanation for why differences noted between hetero-
sexual and gay males on a variety of tasks are much larger (and
more striking) than differences found between lesbians and
heterosexual females (including the differences found in the
present study).
Finally, future research should examine the characteristic
array of odorants present in gays and lesbians using the gas
chromatograph and mass spectrometry techniques utilized in
previous studies on the characteristics of odorants present in
heterosexual males and females (Zeng et al., 1991; Zeng, Ley-
den, Spie lman, & Preti, 1996). It is now clear that additional
care should be taken during subject selection in future studies of
axillary odorants: Sexual orientation should be routinely as-
sessed during recruitment.
Acknowledgments—This work was supported by a Social
Sciences and Humanities Research Council of Canada Fel-
lowship (756-2001-0286) to Y. Martins and by institutional
funds from the Monell Chemical Senses Center. We thank Jason
Eades and Paul Wise for their help in preparation of stimulus
materials and computer programming.
REFERENCES
Binson, D., Michaels, S., Stall, R., Coates, T., Gagnon, J., & Catania, J.
(1995). Prevalence and social distribution of men who have sex
with men: United States and urban centers. Journal of Sex Re-
search, 32, 245–254.
Blanchard, R. (2001). Fraternal birth order and the maternal immune
hypothesis of male homosexuality. Hormones and Behavior, 40,
105–114.
Blanchard, R., & Bogaert, A.F. (1996). Homosexuality in men and
number of older brothers. American Journal of Psychiatry, 153,
27–31.
Diamond, L. (1998). Development of sexual orientation among ado-
lescent and young adult women. Developmental Psychology, 34,
1085–1095.
Ellis, L., & Ames, M.A. (1987). Neurohormonal functioning and sexual
orientation: A theory of homosexuality/heterosexuality. Psycho-
logical Bulletin, 101, 233–258.
Geschwind, N., & Galaburda, A.M. (1985). Cerebral lateralization:
Biological mechanisms, associations, and pathology: II. A hy-
pothesis and a program for research. Archives of Neurology, 42,
521–552.
Green, B.G., Shaffer, G.S., & Gilmore, M.M. (1993). Derivation
and evaluation of a semantic scale of oral sensation magnitude
with apparent ratio properties. Chemical Senses, 18, 683–702.
Henderson, A.F. (1984). Homosexuality in the college years: Devel-
opmental differences between men and women. Journal of Amer-
ican College Health, 32, 216–219.
Ivanyi, P. (1978). Some aspects of the H-2 system, the major histo-
compatibility system in the mouse. Proceedings of the Royal So-
ciety of London, 202, 117–158.
Jacob, S., McClintock, M.K., Zelano, B., & Ober, C. (2002). Paternally
inherited HLA alleles are associated with women’s choice of male
odor. Nature Genetics, 30, 175–179.
Karlson, P., & Luscher, M. (1959). ‘‘Pheromones’’: A new term for a
class of biologically active substances. Nature , 183, 55–56.
Kinsey, A.C., Pomeroy, W.B., & Martin, C.E. (1948). Sexual behavior in
the human male. Philadelphia: W.B. Saunders.
Kitzinger, C., & Wilkinson, S. (1995). Transitions from heterosexuality
to lesbianism: The discursive production of lesbian identities.
Developmental Psychology, 31, 95–104.
Laumann, E., Gagnon, J.H., Michael, R.T., & Michaels, S. (1994). The
social organization of sexuality: Sexual practices in the United
States. Chicago: University of Chicago Press.
Mustanski, B.S., Chivers, M.L., & Bailey, J.M. (2002). A critical review
of recent biological research on human sexual orientation. Annual
Review of Sex Research, 13, 89–140.
Mustanski, B.S., Dupree, M.G., Nievergelt, C.M., Bocklandt, S.,
Schork, N.J., & Hamer, D.H. (2005). A genomewide scan of male
sexual orientation. Human Genetics, 116
, 272–278.
Nyby, J. (1982). Volatile and nonvolatile chemosignals of female ro-
dents: Differences in hormonal regulation. In D. Mu
¨
ller-Schwarze
& R.N. Silverstein (Eds.), Chemical signals in vertebrates 3 (pp.
179–193). New York: Plenum Press.
Ober, C., Weitkamp, L.R., & Cox, N. (1999). HLA and mate choice. In
R.E. Johnston, D. Mu
¨
ller-Schwarze, & P. Sorenson (Eds.), Ad-
vances in chemical signals in vertebrates 8 (pp. 189–199). New
York: Plenum Press.
Ober, C., Weitkamp, L.R, Cox, N., Dytch, H., Kostyu, D., & Elias, S.
(1997). HLA and mate choice in humans. American Journal of
Human Genetics, 61, 497–504.
Porter, R.H., Balogh, R.D., Cernoch, J.M., & Franchi, C. (1986).
Recognition of kin through characteristic body odors. Chemical
Senses, 11, 389–395.
Porter, R.H., & Winberg, J. (1999). Unique salience of maternal breast
odors for newborn infants. Neuroscience and Biobehavioral Re-
views, 23, 439–449.
Preti, G., Cutler, W.B., Huggins, G.R., Garcia, C.R., & Lawley, H.J.
(1986). Human axillary secretions influence women’s menstrual
cycles: The role of donor extracts of females. Hormones and Be-
havior, 20, 474–482.
Preti, G., & Wysocki, C.J. (1999). Human pheromones: Releasers or
primers—fact or myth. In R.E. Johnston, D. Mu
¨
ller-Schwarze, &
P. Sorenson (Eds.), Advances in chemical signals in vertebrates 8
(pp. 315–331). New York: Plenum Press.
Preti, G., Wysocki, C.J., Barnhart, K.T., Sondheimer, S.J., & Leyden,
J.J. (2003). Male axillary extracts contain pheromones that affect
pulsatile secretion of luteinizing hormone and mood in female
recipients. Biology of Reproduction, 68, 2107–2113.
Rasmussen, L.E., Lee, T.D., Roelofs, W.L., Zhang, A., & Daves, G.D.,
Jr. (1996). Insect pheromone in elephants. Nature, 379, 684.
Russell, M.J. (1976). Human olfactory communication. Nature, 260,
520–522.
700 Volume 16—Number 9
Preference for Human Body Odors
Rust, P.C. (1993). ‘‘Coming out’’ in the age of social constructionism:
Sexual identity formation among lesbian and bisexual women.
Gender and Society, 24, 366–386.
Savic, I., Berglund, H., & Lindstro
¨
m, P. (2005). Brain response to pu-
tative pheromones in homosexual men. Proceedings of the Na-
tional Academy of Sciences, USA, 102, 7356–7361.
Sell, R.L., Wells, J.A., & Wypij, D. (1995). The prevalence of homo-
sexual behavior and attraction in the United States, the United
Kingdom and France: Results of national population-based sam-
ples. Archives of Sexual Behavior, 24, 235–248.
Singh, P.B. (2001). Chemosensation and genetic individuality. Repro-
duction, 121, 529–539.
Vandenbergh, J.G. (1989). Coordination of social signals and ovarian
function during sexual development. Journal of Animal Science,
67, 1841–1847.
Weitkamp, L.R., & Ober, C. (1999). Ancestral and recombinant 16-
locus HLA haplotypes in the Hutterites. Immunogenetics, 49,
491–497.
Wyatt, T. (2003). Pheromones and animal behaviour: Communication
by smell and taste. Cambridge, England: Cambridge University
Press.
Yamazaki, K., Singer, A., & Beauchamp, G.K. (1998–1999). Origin,
functions and chemistry of H-2 regulated odorants. Genetica, 104,
235–240.
Yamazaki, K., Singer, A.G., Curran, M., & Beauchamp, G.K. (1999).
Origin, functions and chemistry of H-2 regulated odorants. In R.E.
Johnston, D. Mu
¨
ller-Schwarze, & P. Sorenson (Eds.), Advances
in chemical signals in vertebrates 8 (pp. 173–180). New York:
Plenum Press.
Zeng, X.-N., Leyden, J.J., Lawley, H.J., Sawano, K., Nohara, I., & Preti,
G. (1991). Analysis of the characteristic odors from human male
axillae. Journal of Chemical Ecology, 17, 1469–1492.
Zeng, X.-N., Leyden, J.J., Spielman, A.I., & Preti, G. (1996). Analysis
of the characteristic human female axillary odors: Qualitative
comparison to males. Journal of Chemical Ecology, 22, 237–257.
(R
ECEIVED 12/2/03; REVISION ACCEPTED 1/4/05;
F
INAL MATERIALS RECEIVED 2/1/05)
Volume 16—Number 9 701
Y. Martins et al.
