Content uploaded by Kenneth J. Zucker
Author content
All content in this area was uploaded by Kenneth J. Zucker on Sep 23, 2017
Content may be subject to copyright.
SPECIAL SECTION: THE PUZZLE OF SEXUAL ORIENTATION
Gay Male Only-Children: Evidence for Low Birth Weight and High
Maternal Miscarriage Rates
Malvina N. Skorska
1
•Ray Blanchard
2
•Doug P. VanderLaan
3,4
•Kenneth J. Zucker
2
•
Anthony F. Bogaert
1,5
Received: 8 January 2016 / Revised: 24 May 2016 / Accepted: 2 August 2016 / Published online: 22 August 2016
Springer Science+Business Media New York 2016
Abstract Recent findings suggest that there may be a maternal
immune response underpinning the etiology of sexual orienta-
tion of gay male only-children. This maternal immune response
appears to be distinct from that which is purported to explain the
classic fraternal birth order effect found in studies of male sexual
orientation. We tested two predictions related to the hypothesized
maternal immune response in mothers of gay male only-children:
(1) elevated fetal loss among mothers who have had gay male
only-children and (2) lower birth weight in gay male only-chil-
dren. Mothers of at least one gay son (n=54) and mothers of
heterosexual son(s) (n=72) self-reported their pregnancy histo-
ries, including the birth weights of newborns and number of fetal
losses (e.g., miscarriages). Mothers of gay male only-children
(n=8) reported significantly greater fetal loss compared with
mothers of males with four other sibship compositions (gay with
no older brothers, gay with older brothers, heterosexual only-chil-
dren, heterosexual with siblings) (n=118). Also, firstborn gay
male only-children (n=4) had a significantly lower birth weight
than firstborn children in the four other sibship compositions (n=
59). Duration of pregnancy was not significantly different among
the groups of firstborn children in the birth weight analyses. Thus,
this study found further support for a distinct pattern of maternal
immune response implicated in the etiology of male sexual ori-
entation. Mechanisms that may underlie this potential second type
of maternal immune response are discussed.
Keywords Sexual orientation Birth weight Fetal loss
Miscarriage Birth order Maternal immune hypothesis
Introduction
The study of the biological origins of sexual orientation has
been ongoing for several decades now. Researchers investi-
gating sexual orientation have found evidence for the fol-
lowing biological correlates: genetics (e.g., Bailey, Dunne &
Martin, 2000; Bailey & Pillard, 1991; Hamer, Hu, Magnuson,
Hu, & Pattatucci, 1993; Mustanski et al., 2005; Sanders et al.,
2015), sex-dimorphic brain structures (e.g., Abe
´, Johansson,
Allze
´n, & Savic, 2014; LeVay, 1991; Witelson et al., 2008),
and hormones, particularly at the prenatal level (e.g., Bao &
Swaab, 2011; Ellis & Ames, 1987; Grimbos, Dawood, Burriss,
Zucker, & Puts, 2010;Hines,2011;Lalumie
`re, Blanchard, &
Zucker, 2000; Ngun, Ghahramani, Sanchez, Bocklandt, & Vilain,
2011). Whereas most of these findings apply to both men and
women, a different biological mechanism has been purported to
explain the fraternal birth order (FBO) effect that has been
found in studies of men’s sexual orientation.
In the FBO effect, the odds that men, but not women, will
be same-sex attracted as adults are increased with a greater
number of older brothers (for reviews, see Blanchard,
1997,2004,2008; Blanchard & VanderLaan, 2015; Bogaert
& Skorska, 2011; VanderLaan, Blanchard, Wood, & Zucker,
&Anthony F. Bogaert
tbogaert@brocku.ca
1
Department of Psychology, Brock University, 1812 Sir Isaac
Brock Way, St. Catharines, ON L2S 3A1, Canada
2
Department of Psychiatry, University of Toronto, 250 College
Street, Toronto, ON M5T 1R8, Canada
3
Department of Psychology, University of Toronto Mississauga,
3359 Mississauga Road N., Mississauga, ON L5L 1C6, Canada
4
Child, Youth and Family Division, Underserved Populations
Research Program, Centre for Addiction and Mental Health,
Toronto, ON, Canada
5
Department of Health Sciences, Brock University, 1812 Sir
Isaac Brock Way, St. Catharines, ON L2S 3A1, Canada
123
Arch Sex Behav (2017) 46:205–215
DOI 10.1007/s10508-016-0829-9
2014). A maternal immune mechanism is a hypothesized bio-
logical explanation for this effect. The hypothesis postulates
that with each successive male fetus the mother’s immune
system is repeatedly exposed (and thus potentially immu-
nized) to male-specific proteins associated with the Y-chro-
mosome, which a mother does not have. For a mother who has
been immunized, an increasing concentration of antibodies
develops after each male fetus and crosses the blood–brain
barrier to affect specific areas in the brain associated with the
development of sexual orientation, which would then influ-
ence the sexual orientation of later born sons (Blanchard &
Bogaert, 1996; Blanchard & Klassen, 1997). Recently, Blan-
chard (2012b) proposed a second maternal immune-based
explanation of the etiology of male sexual orientation that is
separate from the one related to the fraternal birth order effect.
Blanchard (2012b) found that in a sample of 44,981 hetero-
sexual and gay firstborn men, gay firstborns hadsignificantly
fewer younger siblings than heterosexual firstborns. As a possi-
ble explanation for this finding of fewer younger siblings in gay
firstborn men, Blanchard hypothesized that there might be a
maternal immune effect that increases the odds of a gay sexual ori-
entation in a subpopulation of firstborn sons and is associated with
a higher likelihood of fetalloss during other pregnancies. Thus,
this maternal immune effect would be at least partially distinct
from the classic FBO effect. This maternal immune effect should
also be more common among the subpopulation of mothers who
have had gay male only-children. The logic supporting the selec-
tion of this specific subpopulation of sons stems from studies tha t
suggest there are links between a maternal immune re sponse to a
fetus, a lower birth weight in a newborn, and secondary recurrent
miscarriage. Specifically, a maternal immune response to a fetus
can reduce the birth weight of a newborn and can prevent future
pregnancies from reaching full term, rendering the newborn an
only-child. Blanchard hypothesized that in addition to affecting
maternal fetal loss and the birth weight of the newborn, the indi-
vidual’s sexual orientation may be affected as well. To support
the logic behind this hypothesis, research literature supporting
the associations between a maternal immune response and fetal
loss andbetween a maternal immune response and birth weight,
areoutlinednext,andfollowedbysomerecentresearchon only-
children, sexual orientation, and birth weight.
A link between fetal loss and a maternal immune effect has
been posited in the scientific literature. There is evidence that
miscarriages in mothers partially reflect their immune response
to a male-specific protein—SMCY—important in male fetal
development (Nielson, 2011). SMCY has a relatively general-
ized role in the development of many structures of the body aside
from the brain during sexual differentiation. In sum, elevated
fetal loss may serve as a marker of a maternal immune response
(e.g., to SMCY or other male-specific proteins).
In addition to fetal loss, a second potential marker of a
maternal immune response is birth weight. An association
between low birth weight of offspring and a mother’s immune
system has also been advanced in the scientific literature. For
example, research in rodent models has demonstrated that male
pups had lower birth weight after maternal immunization to
male-specific antigens (Kahn & Baltimore, 2010). In human
studies, lower birth weight in both male and female offspring
was associated with a mother’s immune system activation during
pregnancy (Christensen et al., 2012; Kiefte-de Jong et al., 2013;
Kusanovic et al., 2007;Milns&Gardner,1989; Nielsen et al.,
2010; Silver et al., 2011) or when the mother and fetus show
antigen incompatibility (e.g., incompatible blood groups between
the mother and fetus) (Hoff, Peevy, Spinnato, Giattina, & Peterso n,
1993). These studies, however, have not investigated the associ-
ation among birth weight, sexual orientation, and sibship status.
There are two published studies examining birth weight in rela-
tion to sexual orientation and whether an individual was an only-
child or an oldest child (i.e., a firstborn with one or more younger
siblings).
Blanchard (2012a) demonstrated a significant interaction
between sexual orientation and sibship status (i.e., only-child
versus oldest child), such that the seven lesbian females and six
gay males who were only-children had a lower birth weight than
the 134 heterosexual females and 74 heterosexual males (and
there was no significant difference in birth weight between the
heterosexual and gay/lesbian oldest children). The mean differ-
ence in birth weight was 241.90 g (d=0.53). This pattern was
consistent with the hypothesis of a separate etiology for gay
male only-children, but both sexes were represented in these
analyses, and thus this study suggests that the separate maternal
immune etiology may not be specific to males.
VanderLaan et al. (2015) reported on a sample of 1722 male
and female children and adolescents, some of whom were clin-
ically referred for gender dysphoria andothers who were clinical
controls. Specifically, there were 1536 children (351 females,
1185 males) and 173 adolescents (72 females, 114 males). For
the children in their sample, the cross-gender behavior associ-
ated with gender dysphoria was treated as an indicator that sex-
ual attraction to the same natal sex was a probable adulthood
sexual orientation outcome among children (Green, 1987; Singh,
2012; Steensma, McGuire, Kreukels, Beekman, & Cohen-Ket-
tenis, 2013; Wallien & Cohen-Kettenis, 2008). For the adoles-
cents, sexual orientation was classified based on questionnaire
responses regarding sexual behavior and attraction. In this sam-
ple, the male only-children in the gender dysphoria group (n=
65) showed relatively lower birth weight compared with the
male only-children in the control group (n=167) (p=.07, d=
0.27 in males; p=.81, d=-0.06 in females). Thus, a similar
pattern emerged as in Blanchard (2012a), but the pattern in Van-
derLaan et al. (2015) was specific to males.
The goal of the current study was to replicate the birth weight
finding in males, given that not all findings for males have been
statistically significant. We focused only on natal male births
206 Arch Sex Behav (2017) 46:205–215
123
because our data set did not include information about lesbian
female births. Also, we conducted the first test of the fetal loss
prediction. Specifically, the current study employed a new data
set to explore whether gay male only-children have a lower birth
weightthan male childrenwith othersibship compositions. Also,
we examined whether mothers of gay male only-children have
experienced greater fetal loss (e.g., miscarriages) than mothers
of male children with other sibship compositions. We predicted
that (1) mothers of gay male only-children would have more fetal
loss than mothers of male children with other sibship composi-
tions and (2) gay male only-children would have a lower birth
weight than male children with other sibship compositions.
Method
Participants
A total of 159 individuals (12 men, 147 women) participated in a
larger study examining the association between a mother’s his-
tory of immunization to male-specific proteins (as inferred from
blood analysis) and her children’s sexual orientations. Partici-
pants were recruited via posters placed around an Ontario uni-
versity campus, advertisements placed on Kijiji, booths set up at
local Pride festivals, ads placed in local LGBT magazines and
radio stations, and ads placed in local newspapers. Ads were
targetedforthe specific samples requiredfor the larger study.
Specifically, ads asked that any interested‘‘mothers of son(s),’’
‘‘mothers of gay son(s),’’‘‘mothers of straight/heterosexual son(s),’’
‘‘mothers of daughters only,’’and‘‘men only’’contact the laboratory
for more information about the study via e-mail or phone. Also,
six participants were mothers of males clinically referred to a spe-
cialty child and adolescent gender identity service for gender
dysphoria.
Of the 159 participants, 12 were men and 147 were women.
The women were further classified based on the sex and sexual
orientationof their offspring. Thus,59 were classifiedas mothers
of heterosexual son(s) only (i.e., could have daughters, but sons
were heterosexual only), 48 were classified as mothers of at least
one gay son (i.e., could have more than one gay son, could have
heterosexual sons as well, could have daughters), 11 were clas-
sified as mothers of daughters only(i.e.,nosons),sixwereclas-
sified as mothers of at least one natal male child clinically refer-
red for gender dysphoria (i.e., could have more than one natal
malechild who experiencedgender dysphoria, couldhave gay or
heterosexual sons, could have daughters), 13 were classified as
mothers of sons who had an unknown sexual orientation because
their sons were too young to know their sexual orientations (i.e.,
couldhave daughters aswell), threewere classifiedas mothersof
at least one transsexual individual, two were classified as moth-
ers of at least one bisexual son, andfivewereclassifiedaswomen
who had no known pregnancies.
For the current study, a subsample of the recruited partici-
pantswas of interest.Specifically,wewereinterestedinthemain
groups from the larger study: mothers of at least one gay son and
mothersof heterosexual son(s)only. Thus, the mothersof hetero-
sexual son(s) only (n=59), mothers of at least one gay son (n=
48), mothers of at least one natal male child clinically referred for
gender dysphoria (n=6), and the mothers of sons who had an
unknown sexual orientation (n=13) were included in the cur-
rent study, for a total of 126 mothers. The mothers of at least one
natal male child clinically referred for gender dysphoria were re-
classified as mothers of at least one gay son becaus e most of these
gender dysphoric sons were likely to be gay men as adults (Green,
1987; Singh, 2012;Steensmaetal.,2013; Wallien & Cohen-
Kettenis, 2008). In total, then, 54 mothers had at least one gay
son. Similarly, the mothers of sonswho had an unknown sexual
orientation were re-classified as mothers of heterosexual son(s)
only, because 95–98 % of these children were likely to be hetero-
sexual as adults (Laumann, Gagnon, Michael, & Michaels, 1994;
LeVay, 2010). In total, then, 72 mothers were mothers of hetero-
sexual son(s) only. The majority of the 126 mothers were Cau-
casian (n=99 of the n=118 who completed the ethnicity ques-
tion; 78.6 %), and were attending or had completed a community
college diploma, university degree, or higher (n=94; 75 %).
Theiragesrangedfrom27to78years(M=51.30, SD =10.72).
Measures
Only the measures of interest to the current study are described
below.
1
Fetal Loss
Each mother was asked to report the outcome of her preg-
nancies. Outcomes were labeled by a mother as a miscarriage/
abortion, stillbirth, or live birth. For each mother, the number
of total miscarriages/abortions and stillbirths was tabulated
across all pregnancies to represent the fetal loss variable.
Thus, fetal loss represents the total number of fetuses/preg-
nancies that did not result in a live birth across all known
fetuses/pregnancies. In the rare instances of twins or multiple
fetuses for one pregnancy, each fetus was counted as one preg-
nancy and/or lost fetus. See‘‘Appendix’’ for the form each
mother was asked to complete for each pregnancy.
Duration of Pregnancy
Each mother was asked to report the duration of a pregnancy
in weeks for pregnancies with all outcomes.
1
Please contact the corresponding author for a list of all measures.
Arch Sex Behav (2017) 46:205–215 207
123
Birth Weight
Each mother was asked to report the birth weight of each fetus in
pounds and ounces, grams, or else indicate that the birth weight
was unknown (e.g., if the fetus resulted in a miscarriage). Birth
weights reported in pounds and ounces were converted to grams.
Sex of Offspring
Each mother was asked to report the sex of each fetus using the
designation male, female, or unknown.
Sexual Orientation of Offspring
Each mother was asked to report the sexual orientation of each
live-born child using the labels heterosexual, gay, lesbian, bisex-
ual, transsexual,
2
or unknown. Mothers of natal male children
clinically referred for gender dysphoria did not answer a ques-
tion about the sexual orientation of their offspring. As indicated
above, most natal male children clinically referred for gender
dysphoria are likely to be homosexual as adults, with or without
continuing gender dysphoria.
Age of Mothers at Participation
Each mother was asked to report her date of birth. Age was cal-
culatedas the dateof examinationminus reporteddate of birth,in
years.
Procedure
A research assistant and phlebotomist met the participant at their
homeor on campus(whichever was mostconvenient). After pro-
viding consent, the participant completed a demographics and
pregnancy history questionnaire. Along with the measures of
reproductive history mentioned above, this questionnaire con-
tained questions related to demographic characteristics of the
participant (e.g., ethnicity, education). After completion of the
questionnaire, a sample of blood was drawn from the partici-
pant. Then, the participant was thanked, debriefed, and compen-
sated monetarily for their participation. This study was approved
by the necessary Research Ethics Boards.
Statistical Analyses
Fetal Loss Analyses
The 126 mothers were first re-classified into five groups based
on their pregnancy historie s: 8 mothers of gay male only-children,
23 mothers of gay males with no older brothers,
3
23 mothers of
gay males with older brothers, 11 mothers of heterosexual male
only-children, and 61 mothers of heterosexual males with sib-
lings. Analysis of variance (ANOVA) was used to test whether
there were any age differences among the five groups of mothers.
Helmert contrasts were used for planned comparisons to deter-
mine whether there were any differences among the five groups of
mothers (independent variable) in total number of fetuses lost
(dependent variable).
For five groups, Helmert contrasts compare Group 1 with the
mean of all later groups (i.e., 2, 3, 4, and 5); Group 2 with the
mean of all later groups (i.e., 3, 4, and 5); Group 3 with the mean
of all later groups (i.e., 4 and 5); and Group 4 with the mean of all
later groups (i.e., 5). The first three groups comprise the mothers
of gay males and the latter two groups comprise the mothers of
heterosexual males. Group 1 is designated as mothers of gay male
only-childrenbecause our predictions (basedon prior research and
theory) concern this group having the greatest number of fetal
losses. Group 2 is mothers of gay males with no older brothers
because their total reproductive output seems more similar to that
of Group 1 mothers compared with Group 3 mothers. Group 3 is
mothers of gay males with older brothers because that is the last
group of mothers within‘‘mothers of gay males.’’Group 4 is moth-
ers of heterosexual male only-children because this group seems
more similar to Group 1 in terms of total reproductive output than
Group 5. Group 5 is mothers of heterosexual males with siblings
because that is the remaining group. Omnibus ANOVA was not
performed because we had planned comparisons, and with plan-
nedcomparisons we canmove directly tothe comparisons(Tabach-
nick & Fidell, 2007,p.52).
A potential confound was that mothers would be expected to
have more fetal loss if they are likely to be pregnant more often
and thus have more children. Thus, as a second dependent vari-
able, a ratio of the number of lost fetuses to the number of live
births was computed.
4
Again, Helmert contrasts were used for
planned comparisons with the five groups of mothers as the inde-
pendent variable and the ratio variable as the dependent variable.
The same group ordering was utilized here as in the previous
analysis using Helmert contrasts.
Birth Weight Analyses
In the birth weight analyses, the units of analysis were selected
sons rather than the mothers. Because these sons did not actu-
ally participate in the study, they are referred to as subjects to
avoid confusion.
2
Although‘‘transsexual’’does not refer to sexualorientation,we included
thisdesignationundersexual orientationin the questionnaireto give moth-
ers another option if one or more of their children had a trans identity.
3
One of these mothers indicated that her son was gay/bisexual.
4
We decided to keep both analyses because the ratio variable is subject
to its own limitations (e.g., deviations from non-normality). Thus, we
believe a more complete picture can be gleaned from the data by
analyzing the fetal loss variable using both non-ratio and ratio variables.
208 Arch Sex Behav (2017) 46:205–215
123
In these birth weight analyses, we addressed a different
potential confound: Birth weight tends to increase over suc ceed-
ing pregnancies (Wilcox,Chang, & Johnson, 1996).To control
for this possible relation, we restricted the analysis to live-born
sons who were the result of their mother’s first known pregnancy
(n=63). In other words, we selected subjects who were identical
on maternal gravidity (number of pregnancies) and maternal
parity (number of deliveries) for their mother’s first pregnancy,
which is a stricter criterion than controlling for maternal parity
alone. Thus, a subset of the sample used to examine fetal loss
was used fo r the birth weigh t analyses because notall subjects
in the fetal loss data were firstborns.
For these birth weight analyses, it is easierto label the sons
according to their sibship composition rather than their
mothers’ reproductive histories, although there is a one-to-one
correspondence between the two frames of reference, as dis-
cussed below (see also Table 1). The sons were categorized as
follows: 4 gay male only-children, 7 gay males with no older
brothers, 14 heterosexual males with gay younger brothers, 10
heterosexual male only-children, and 28 heterosexual males
with siblings (other than gay younger brothers).
It is important to note the correspondence between the grou ps
of mothers studied in the analyses of fetal loss and the groups of
offspring studied in the analyses of birth weight (see Table 1).
This correspondence is immediately apparent in four of the five
instances (e.g., for the mothers of gay only-children, the off-
spring were gay only-children). The exception is the mothers of
gay males with older brothers, whose first-pregnancy offspring
were heterosexual males with gay younger brothers.
ANOVA was used to test whether there were any age dif-
ferences among the mothers of the five groups of subjects.
Helmert contrasts were used in planned comparisons to deter-
mine whether there were any differences among the five groups
of subjects (independent variable) in birth weight (dependent
variable). Groups were ordered the same way as their counter-
parts in the fetal loss analyses because there was no reason to
predict a different group ordering. As with the fetal loss analyses,
due to having planned comparisons, omnibus ANOVA was not
performed (Tabachnick & Fidell, 2007).
Results
Fetal Loss Analyses
The ANOVA with age at examination as the dependent vari-
able was marginally significant, F(4, 121) =2.43, p=.05.
Planned comparisons were not conducted because there were
no a priori expectations of age differences among the groups
of mothers used in the fetal loss analyses.
Of the 126 mothers in the sample, 40 reported a loss of 62
fetuses. All but two of these losses were reported as a miscar-
riage; in one of the remaining cases, the information was missing
and in the other case, the pregnancy resulted in a stillbirth.
Helmert contrasts revealed that a significant difference was
found between the mean of the Group 1 mothers compared
with the mean of all other groups (M
diff
=.84, SE =.36,
p=.02, 95 % CI 0.11, 1.56) (M
Group1
=1.25, SD
Group1
=1.39;
M
Group2
=0.43, SD
Group2
=0.79; M
Group3
=0.74, SD
Group3
=
1.71; M
Group4
=0.09, SD
Group4
=0.30; M
Group5
=0.39,
SD
Group5
=0.64). That is, the mothers of gay male only-children
had significantly greater numbers of fetal losses compared with
the mean of all other mothers (d=0.76; see Fig. 1). None of the
remaining Helmert contrasts was significant.
With ratio of lost fetuses to live births as the dependent
variable, Helmert contrasts revealed a significant difference
between the mean of the Group 1 mothers compared with
the mean of all other groups (M
diff
=1.09, SE =.18, p\
.001, 95 % CI 0.74, 1.44) (M
Group1
=1.25, SD
Group1
=1.39;
M
Group2
=0.14, SD
Group2
=0.26; M
Group3
=0.24, SD
Group3
=0.56;
M
Group4
=0.09, SD
Group4
=0.30; M
Group5
=0.17, SD
Group5
=
0.28). Thus, the mothers of gay male only-children had a
significantly greater mean ratio of the number of fetal losses
to number of live births compared with the mean of all other
mothers (d=1.55; see Fig. 2). Again, none of the remaining
Helmert contrasts was significant.
Given the typical non-normal distribution of ratio variables,
we also examined the ratio data using nonparametric statistics.
The most appropriate analysis is the Kruskal–Wallis test, which
is the nonparametric equivalent of an ANOVA (Field, 2013).
Themean ranks of theratio variable differedsignificantly among
the five groups of mothers, H=9.78, df =4, p=.04. There are
two possible ways to follow-up the Kruskal–Wallis test in SPSS
if the omnibus is significant (Field, 2013). We chose the‘‘step-
wise step-down’’option, which demonstrated that the mothers of
gay only-children were driving the significant omnibus H. Specif-
ically, the mothers of gay only-children were a separategroup
from the four other groups of mothers (mothers of gay males
with no older brothers, mothers of gay males with older brothers,
mothers of heterosexual male only-children, mothers of hetero-
sexual males with siblings) based on their mean ranks. The four
other groups of mothers, however, did not differ significantly
fromone another interms ofmean ranks (p=.58).Thus, themoth-
ers of gay only-children had a significantly greater mean rank of
the ratio of the number of fetal losses to number of live births com-
pared with the mean rank of all other mothers, convergent with the
results of the parametric test.
Birth Weight Analyses
TheANOVA with mother’sage atexamination as thedependent
variable was not significant, F(4, 58) =2.19, p=.08. Planned
comparisons were not conducted because there were no a priori
expectations of age differences among the groups of mothers
relevant for the birth weight analyses.
Arch Sex Behav (2017) 46:205–215 209
123
Planned comparisons using Helmert contrasts showed that
the predicted birth weight difference between the mean of
Group 1 subjects compared with the mean of all other groups
of subjects was significant (M
diff
=-615.73, SE =259.02,
p=.021, 95 % CI -1134.21, -97.25) (M
Group1
=2970.00,
SD
Group1
=623.48; M
Group2
=3713.71, SD
Group2
=488.28;
M
Group3
=3489.07, SD
Group3
=477.20; M
Group4
=3506.80,
SD
Group4
=411.93; M
Group5
=3633.32, SD
Group5
=517.41).
Thus, the mean birth weight of gay male only-children was
significantly lower than the mean birth weight of the first-preg-
nancy males in all other groups (Fig. 3). None of the remaining
Helmert contrasts was significant. In absolute terms, the differ-
ence in birth weight was approximately 600 g, representing a
large effect (d=1.21).
Additional Analyses Related to Birth Weight
We conducted a follow-up analysis to rule out another
explanation for the birth weight effect. Duration of pregnancy
did not differ among the same groups of subjects included in
the previous analyses, F(4, 58) =.38, p=.83. Thus, mean
duration of pregnancy of the first-pregnancy sons was virtually
identical for all groups of subjects (and close to the typical
40 weeks).
Discussion
In the current study, we found that mothers of gay male only-
children reported, on average, significantly greater mean fetal
loss compared with the mean of mothers of gay males with no
older brothers, mothers of gay males with older brothers, moth-
ers of heterosexual male only-children, and mothers of hetero-
sexual males with siblings, supporting Prediction 1. Notably, the
fetal loss effect was particularly evident for the ratio measure-
ment (fetal loss/live births), an arguably better measure than mean
absolutefetal loss, as the formeraccounts for overallnumber of
pregnancies. Further, we found that first-gestated gay male
Table 1 Correspondence between the groups included in the fetal loss analyses and the groups included in the birth weight analyses
Fetal loss analyses Birth weight analyses
Group name nPregnancy history Group name n
1. Mothers of gay male only-children 8 G 1. Gay male only-children 4
2. Mothers of gay males with no older brothers 23 e.g., GHH 2. Gay males with no older brothers 7
3. Mothers of gay males with older brothers 23 e.g., HHG 3. Heterosexual males with gay younger brothers 14
4. Mothers of heterosexual male only-children 11 H 4. Heterosexual male only-children 10
5. Mothers of heterosexual males with siblings 61 e.g., HH 5. Heterosexual males with siblings 28
Dependent variables
Number of fetuses lost by the mother
due to miscarriage, abortion, and stillbirth
Birth weight of the first-born male only
Ggay son, Hheterosexual son
Fig. 1 Mean (±SEM) number of
lost fetuses for each group of
mothers
210 Arch Sex Behav (2017) 46:205–215
123
only-children had, on average, a significantly lower mean
birth weight than the mean of first-gestated children in each
of the following sibship categories: gay males with no older
brothers, heterosexual males with gay younger brothers, hetero-
sexual male only-children, and heterosexual males with siblings
(other than gay younger brothers), supporting Prediction 2. We
have also ruled out a potential explanation of the duration of preg-
nancy for the birth weight finding. The former finding regarding
fetal loss is novel, whereas the latter finding regarding birth weight
is a partial replication and extension of the findings of Blanchard
(2012a) within males and VanderLaan et al. (2015).
The present study provided further support for Blanchard’s
(2012b) hypothesis of a separate etiology of male sexualorienta-
tion related to a mother’s immune response in gay only-children.
Specifically, two markers of a maternal immune response—high
fetal loss and low birth weight—were evident in gay male only-chil-
Fig. 2 Mean (±SEM) ratio of
lost fetuses to live births for each
group of mothers
Fig. 3 Mean (±SEM) birth
weight for each group of first-
pregnancy sons.‘‘Heterosexual
Males with Gay Younger
Brothers’’are first-pregnancy
sons of mothers of gay males with
older brothers (cf. Figs. 1,2). The
category-label‘‘Heterosexual
Males with Siblings’’means
siblings that do not include a gay
younger brother
Arch Sex Behav (2017) 46:205–215 211
123
dren and in the mothers of these children. Indeed, the effects were
large (e.g., d=1.55 for fetal loss using the ratio measure, and
d=1.21 for birth weight), suggesting that a maternal immune
response may be a particularly powerful agent in the etiology of
gay male only-child individuals andinthefetallossoftheirmoth-
ers. In other words, low birth weight and high fetal loss may be
markers of a powerful form of a maternal immune response asso-
ciated with homosexuality in men, but one that is particularly
found in gay male only-children compared with male children of
other sibship compositions. This type of immune response may be
most detectable in gay male only-children because the absence of
additional siblings in gay male only-children is a sign that many of
their mothers may have been characterized by this powerful
immune response resulting in elevated fetal loss (and hence
no other children).
Regarding lower birth weight among gay male only-children,
the effect size found in the current study (d=1.21) was larger
compared with the effect size of d=0.53 found in Blanchard
(2012a)andd=0.27 found in VanderLaan et al. (2015). These
differences in effect size across studies may be due to the diver-
sity of the samples, the different and various sexual orientation
measures utilized in the studies, or the various sibship composi-
tion breakdowns in the studies. Further research with larger sam-
ple sizes and detailed sibship compositions will help to clarify
the effect size related to the birth weight of gay only-children.
We can only speculate about the specific maternal immune
mechanisms that underlie this potential second type of maternal
immune response. VanderLaan et al. (2015) postulated that moth-
ers who carry the HLA class II allele, HYrHLA, and a homozy-
gous 14 base pair in exon 8 of the HLA-Ggene would be more
likely to have a gay son. This hypothesis was proposed because
of previous literature indicating HYrHLA increased a mother’s
immune response to male-specific minor HY antigens; HYrHLA
was associated with low birth weight in firstborn sons (not first-
born daughters), and the homozygous base pair insertion in exon
8oftheHLA-Ggene was associated with greater miscarriages
after the birth of the first child (Christensen et al., 2012). The inves-
tigation of whether this specific genotype in mothers can explain the
putativesecond type of maternalimmune responsefound in the
current study, in Blanchard (2012a), and in VanderLaan et al.
(2015) will have to await future research.
There may also be different male-specific proteins asso-
ciated with the classic FBO maternal immune effect and this
second potential maternal immune effect associated with gay
only-children. As mentioned in the Introduction, SMCY is
reported to underlie fetal miscarriages (Nielson, 2011) and
thus this male-specific protein may play a role in the effects
associated with this second potential maternal immune
response in gay only-children. During sexual differentiation,
SMCY has a relatively generalized role in the development of
many structures of the body aside from the brain and thus
maternal anti-SMCY (e.g., antibodies to SMCY) likely has
dramatic effects on the fetus (i.e., miscarriage). If so, mothers
of gay only-children may be characterized by a powerful
immune response that includes reactions to a number of male-
specific proteins (including SMCY), which have particularly
powerful effects on additional male fetuses. In contrast, moth-
ers of gay men with older brothers may havehad a weaker but a
more specialized immune response to one or more male-speci-
fic proteins associated with brain development more directly
(e.g., PCDH11Y or NLGN4Y; for reviews, see Blanchard,
2008;Bogaert&Skorska,2011). Again, this speculation waits
further testing.
Limitations and Future Directions
This study cannot address the association between sibship com-
position, sexual orientation, and birth weight in women; as such,
Blanchard’s (2012a) finding that lesbian only-children had lower
birth weights than heterosexual female only-children—and Van-
derLaan et al.’s (2015) lack of such a finding—will require further
investigation. If the second type of maternal immune effect is
indeed replicated in women, the explanations posited in the cur-
rent study may not apply. For example, it may be the case that a
maternal immune response to a non-male-specific protein may
be occurring (Blanchard, 2012a); however, future research is
required to fully elucidate whether the effect occurs in women
and, if so, why. Also, studies utilizing a larger sample size will be
required to replicate and extend the current study. Blanchard’s
(2012a) suggestion of an online survey seems to be a straight-
forward methodological suggestion to managethe sample size
problem.Further,the currentstudyrelied on self-reportsof fetal
loss and birth weight. Although parent-reports of birth weight
seem to be very accurate (e.g., VanderLaan et al. reported r=
.97, p\.001for a subset of participants for whomboth hospital
records and parent-report birth weight were ava ilable), the accu-
racyof self-reports offetal lossmay bemuch lower.For example,
some women may experience a miscarriage without knowing
that one occurred. Hospital records and a longitudinal study
following mothers throughout their childbearing years may help
with determining whether self-reports are reliable indicators of
fetal loss; however, both options are less practical than the self-
reporting of fetal loss by mothers. Nevertheless, future research
incorporating other ways to measure fetal loss would be ben-
eficial.
Conclusion
We found that mothers of gay male only-children experienced
greater mean fetal loss compared with mothers of males with
other sibship compositions. Also, we found that first-gestated
gay male only-children had a lower mean birth weight than first-
gestated male children with other sibship compositions. In sum,
212 Arch Sex Behav (2017) 46:205–215
123
the current study provides additional support for the hypothesis
that a separate etiology of male sexual orientation related to a
mother’s immune response exists—one that is particularly pow-
erful and most detectable in gay male only-children.
Acknowledgments Thank you to J. Bramley, K. Fallis, I. Gabrie, J.
Gabrie, M. Hoffarth, L. Jamieson, K. Kilyk, K. Labanowicz, K. Lee, D.
Mahoney, S. Mazzuocco, S. Norgaard, K. Ross, K. Walczyk, C. Wang,
and N. Wickramasuriya for assistance with data collection. Thank you to
M. Ashton and D. Molnar for assistance with statistical analyses. Thank
you to C. M. McCormick for helpful comments on drafts of this paper. A
version of this paper was included in the Ph.D. dissertation for M. N. Sko-
rska. This research was supported by Natural Sciences and Engineering
Research Council of Canada grant to A. F. Bogaert and R. Blanchard [334-
737-007].
Compliance with ethical standards
Conflict of interest The authors declare no conflict of interest.
Human and animal rights and Informed consent All procedures
performed in studies involving human participants were in accordance
with the ethical standards of the institutional and/or national research
committee and with the 1964 Helsinki Declaration and its later amend-
ments or comparable ethical standards. Informed consent was obtained
from all individual participants included in the study.
Appendix: Questions mothers were asked to answer
regarding each pregnancy they experienced
Pregnancy number X
Your age at end of this pregnancy: _____
Father identifier (i.e., father of the fetus; first name, nickname, initials, number, etc.):_____
Medical problems during pregnancy—circle the answer: Yes No. If yes, please describe below.
Medical problems during delivery—circle the answer: Yes No. If yes, please describe below.
Any method of assisted reproduction (e.g., in vitro fertilization)—circle the answer: Yes No. If
yes, please describe below.
If this pregnancy resulted in one or more live births, does the child/children have (or had) any
medical/health issues—circle the answer: Yes No. If yes, please describe below:
Length of pregnancy (in weeks)_____
If a baby was delivered, method of delivery—circle the answer: C-section Natural.
Number of fetuses (babies carried) in this pregnancy: ____
Here is a table that asks about the outcome(s) of this pregnancy. Note that although most
pregnancies contain one fetus, we are listing up to four fetuses in the event of that you
carried more than one (e.g., twins) in this pregnancy.
Fetus/baby
number
Fetus
identifier
(first
name,
nickname,
initials,
number,
etc.)
Outcome—
miscarriage
or abortion
(MA),
stillbirth (S),
or live birth
(L)
Sex of
fetus—
male (M),
female (F),
or unknown
(U)
Writing
hand—
right (R),
left (L),
both (B),
or
unknown
(U)
Weight at
birth in
grams or
pounds
or
unknown
(U)
Sexual
orientationa—
heterosexual (H),
gay (G), lesbian
(L), bisexual (B),
transsexual (T),
or unknown (U)
1
2
3
4
Did you intentionally try to have any more children after this pregnancy was finished?
___Yes ___No
aNote that mothers of natal male children clinically referred for gender dysphoria were not asked
to answer this question regarding the sexual orientation of each fetus.
Arch Sex Behav (2017) 46:205–215 213
123
References
Abe
´, C., Johansson, E., Allze
´n, E., & Savic, I. (2014). Sexual orientation
related differences in cortical thickness in male individuals. PLoS
ONE, 9, e114721. doi:10.1371/journal.pone.0114721.
Bailey, J. M., Dunne, M. P., & Martin, N. G. (2000). Genetic and envi-
ronmental influences on sexual orientation and its correlates in an
Australian twin sample. Journal of Personality and Social Psychol-
ogy, 78, 524–536. doi:10.1037/0022-3514.78.3.524.
Bailey, J. M., & Pillard, R. C. (1991). A genetic-study of male sexual orien-
tation. Archives of General Psychiatry, 48, 1089–1096. doi:10.1001/
archpsyc.1991.01810360053008.
Bao, A., & Swaab, D. F. (2011). Sexual differentiation of the human brain:
Relation to gender identity, sexual orientation and neuropsychiatric
disorder. Frontiers in Neuroendocrinology, 32, 214–226. doi:10.1016/
j.yfrne.2011.02.007.
Blanchard, R. (1997). Birth order and sibling sex ratio in homosexual versus
heterosexual males and females. Annual Review of Sex Research, 8,
27–67. doi:10.1080/10532528.1997.10559918.
Blanchard, R. (2004). Quantitative and theoretical analyses of the relation
between older brothers and homosexuality in men. Journal of Theoret-
ical Biology, 230, 173–187. doi:10.1016/j.jtbi.2004.04.021.
Blanchard, R. (2008). Review and theory of handedness, birth order, and
homosexuality in men. Laterality, 13, 51–70. doi:10.1080/13576
500701710432.
Blanchard, R. (2012a). A possible second type of maternal–fetal immune
interaction involved in both male and female homosexuality. Archives
of Sexual Behavior, 41,1507–1511. doi:10.1007/s10508-011-9896-0.
Blanchard, R. (2012b). Fertility in the mothers of firstborn homosexual
and heterosexual men. Archives of Sexual Behavior, 41, 551–556.
doi:10.1007/s10508-01-9888-0.
Blanchard, R., & Bogaert, A. F. (1996). Homosexuality in men and number
of older brothers. American Journal of Psychiatry, 153, 27–31. doi:10.
1176/ajp.153.1.27.
Blanchard, R., & Klassen, P. (1997). H-Y antigen and homosexuality in men.
Journal of Theoretical Biology, 185, 373–378. doi:10.1006/jtbi.1996.
0315.
Blanchard, R., & VanderLaan, D. P. (2015). Commentary on Kishida and
Rahman (2015), including a meta-analysis of relevant studies on frater-
nal birth order and sexual orientation in men. Archiv es of Sexual Behav-
ior, 44, 1503–1509. doi:10.1007/s10508-015-0555-8.
Bogaert, A. F., & Skorska, M. (2011). Sexual orientation, fraternal birth order,
and the maternal immune hypothesis: A review. Frontiers in Neuroen-
docrinology, 32, 247–254. doi:10.1016/j.yfrne.2011.02.004.
Christensen,O.B.,Kolte,A.M.,Dahl,M.,Larsen,E.C.,Steffensen,R.,
Nielsen, H. S., …Hviid, T. V. (2012). Maternal homozygosity for a 14
base pair insertion in exon 8 of the HLA-G gene and carriage of HLA
class II alleles restricting HY immunity predispose to unexplained
secondary recurrent miscarriage and low birth weight in children born to
these patients. Human Immunology,73, 699–705. doi:10.1016/j.hum
imm.2012.04.014
Ellis, L., & Ames, M. A. (1987). Neurohormonal functioning and sexual
orientation: A theory of homosexuality-heterosexuality. Psychological
Bulletin, 101, 233–258. doi:10.1037/0033-2909.101.2.233.
Field, A. (2013). Discovering statistics using IBM SPSS statistics (4th ed.).
London: Sage Publications.
Green, R. (1987). The‘‘sissy boy syndrome’’ and the development of homo-
sexuality. New Haven, CT: Yale University Press.
Grimbos, T., Dawood, K., Burriss, R. P., Zucker, K. J., & Puts, D. A.
(2010). Sexual orientation and the second to fourth finger length ratio:
A meta-analysis in men and women. Behavioral Neuroscience, 124,
278–287. doi:10.1037/a0018764.
Hamer, D. H., Hu, S., Magnuson, V. L., Hu, N., & Pattatucci, A. M. L.
(1993). A linkage between DNA markers on the X-chromosome and
male sexual orientation. Science, 261, 321–327. doi:10.1126/science.
8332896.
Hines, M. (2011). Prenatal endocrine influences on sexual orientation and on
sexually differentiated childhood behavior. Frontiers in Neuroendocrinol
ogy, 32, 170–182. doi:10.1016/j.yfrne.2011.02.006.
Hoff, C., Peevy, K. J., Spinnato, J. A., Giattina, K., & Peterson, R. D. A.
(1993). Association between maternal fetal HLA-DR relationships
and fetal growth. American Journal of Reproductive Immunology,
30, 246–253. doi:10.1111/j.1600-0897.1993.tb00626.x.
Kahn, D. A., & Baltimore, D. (2010). Pregnancy induces a fetal antigen-
specific maternal T regulatory cell response that contributes to tolerance.
Proceedings of the National Academy of Sciences of the United States of
America, 107, 9299–9304. doi:10.1073/pnas.1003909107.
Kiefte-de Jong, J. C., Jaddoe, V. W. V., Uitterlinden, A. G., Steegers, E.
A. P., Willemsen, S. P., Hofman, A., …Moll, H. A. (2013). Levels
of antibodies against tissue transglutaminase during pregnancy are
associated with reduced fetal weight and birth weight. Gastroen-
terology,144, 726–735. doi:10.1053/j.gastro.2013.01.003
Kusanovic, J. P., Romero, R., Hassan, S. S., Gotsch, F., Edwin, S.,
Chaiworapongsa, T., …Espinoza, J. (2007).Maternal soluble CD30 is
increased in normal pregnancy, but decreased in preeclampsia and
small for gestational age pregnancies. Journal of Maternal-Fetal and
Neonatal Medicine,20, 867–878. doi:10.1080/14767050701482993
Lalumie
´re, M. L., Blanchard, R., & Zucker, K. J. (2000). Sexual orientation
and handedness in men and women: A meta-analysis. Psychological
Bulletin, 126, 575–592. doi:10.1037//0033-2909.126.4.575.
Laumann, E. O., Gagnon, J. H., Michael, R. T., & Michaels, S. (1994).
The social organization of sexuality: Sexual practices in the United
States. Chicago, IL: University of Chicago Press.
LeVay, S. (1991). A difference in hypothalamic structure between hetero-
sexual and homosexual men. Science, 253, 1034–1037. doi:10.1126/
science.1887219.
LeVay, S. (2010). Gay, straight, and the reason why: The science of
sexual orientation. New York, NY: Oxford University Press.
Milns, N. R., & Gardner, I. D. (1989). Maternal T cells and human pregnancy
outcome. Journal of Reproductive Immunology, 15, 175–178. doi:10.
1016/0165-0378(89)90037-5.
Mustanski,B.S.,DuPree,M.G.,Nievergelt,C.M.,Bocklandt,S.,
Schork, N. J., & Hamer, D. H. (2005). A genome-wide scan of male
sexual orientation. Human Genetics, 116, 272–278. doi:10.1007/
s00439-004-1241-4.
Ngun, T. C., Ghahramani, N., Sanchez, F. J., Bocklandt, S., & Vilain, E.
(2011). The genetics of sex differences in brain and behavior. Frontiers
in Neuroendocrinology, 32, 227–246. doi:10.1016/j.yfrne.2010.10.
001.
Nielsen, H. S. (2011). Secondary recurrent miscarriage and H-Y immunity.
Human Reproduction Update, 17, 558–574. doi:10.1093/humupd/
dmr005.
Nielsen, H. S., Wu, F., Aghai, Z., Steffensen, R., van Halteren, A. G.,
Spierings, E., …Goulmy, E. (2010). HY antibody titers are increased in
unexplained secondaryrecurrentmiscarriagepatients and associated
with low male:female ratio in subsequent live births. Human Repro -
duction,25, 2745–2752. doi:10.1093/humrep/deq242
Sanders,A.R.,Martin,E.R.,Beecham,G.W.,Guo,S.,Dawood,K.,
Rieger, G., …Bailey, J. M. (2015). Genome-wide scan demonstrates
significant linkage for male sexual orientation. Psychological Medicine ,
45, 1379–1388. doi:10.1017/S0033291714002451
Silver, K. L., Conroy, A. L., Leke, R. G. F., Leke, R. J. I., Gwanmesia, P.,
Molyneux, M. E., …Kain, K. C. (2011). Circulating soluble endoglin
levels in pregnant women in Cameroon and Malawi—Associations
with placental malaria and fetal growth restriction. PLoS ONE,6,
e24985. doi:10.1371/journal.pone.0024985
Singh, D. (2012). Afollow-up study of boys with gender identity disorder.
Unpublished doctoral dissertation, University of Toronto.
214 Arch Sex Behav (2017) 46:205–215
123
Steensma, T. D., McGuire, J. K., Kreukels, B. P. C., Beekman, A. J., &
Cohen-Kettenis, P. T. (2013). Factors associated with desistence
and persistence of childhood gender dysphoria: A quantitative fol-
low-up study. Journal of the American Academy of Child and Ado-
lescent Psychiatry, 52, 582–590. doi:10.1016/j.jaac.2013.03.016.
Tabachnick, B. G., & Fidell, L. S. (2007). Using multivariate statistics
(5th ed.). Boston: Pearson Education Inc.
VanderLaan, D. P., Blanchard, R., Wood, H., Garzon, L. C., & Zucker,
K. J. (2015). Birth weight and two possible types of maternal effects
on male sexual orientation: A clinical study of children and ado-
lescents referred to a gender identity service. Developmental Psy-
chobiology, 57, 25–34. doi:10.1002/dev.21254.
VanderLaan, D. P., Blanchard, R., Wood,H., & Zucker, K. J. (2014).Birth
order and sibling sex ratio of children and adolescents referred to a
gender identity service. PLoSONE, 9,e90257.doi:10.1371/journal.
pone.0090257.
Wallien, M. S., & Cohen-Kettenis, P. T. (2008). Psychosexual outcome
of gender-dysphoric children. Journal of the American Academy of
Child and Adolescent Psychiatry, 47, 1413–1423. doi:10.1097/CHI.
0b013e31818956b9.
Wilcox, M. A., Chang, A. M. Z., & Johnson, I. R. (1996). The effects of
parity on birthweight using successive pregnancies. Acta Obste-
tricia et Gynecologica Scandinavica, 75, 459–463. doi:10.3109/
00016349609033354.
Witelson, S. F., Kigar, D. L., Scamvougeras, A., Kideckel, D. M., Buck,
B., Stanchev, P. L., …Black, S. (2008). Corpus callosum anatomy
in right-handed homosexual and heterosexual men. Archives of Sex-
ual Behavior,37, 857–863. doi:10.1007/s10508-007-9276-y
Arch Sex Behav (2017) 46:205–215 215
123
A preview of this full-text is provided by Springer Nature.
Content available from Archives of Sexual Behavior
This content is subject to copyright. Terms and conditions apply.
