hrep$$0320

Human Reproduction vol.12 no.3 pp.611–616, 1997

Length of the follicular phase, time of insemination,

coital rate and the sex of offspring

John F.Martin The present paper extends these ideas and those advanced
recently by the present author (Martin, 1995). The paper argues
Department of Anthropology, Arizona State University, Tempe,
Arizona 85287-2402, USA that, while the penetrability of cervical mucus may improve
as ovulation approaches, debris left in the cervix by previous
The penetrability of cervical mucus improves over the inseminations may impede the passage of spermatozoa insem-
follicular phase. When the length of the follicular phase inated later, thereby dampening any hormonally-induced
varies due to variation in the timing of the luteinizing improvement in mucus penetrability. Y selection may thus be
hormone surge, mucus penetrability will also improve affected by the patterning and frequency of coitus in the fertile
as the phase lengthens. As selection for Y spermatozoa period. It is then suggested that hormonally-induced changes
decreases with improvements in mucus penetrability, sex in mucus penetrability, the effects of the coital pattern and the
ratios at conception should decline in longer follicular superior motility of some Y spermatozoa account for the
phases. Sex ratios should also decline as the time of differences between those studies in which the sex of offspring
insemination approaches ovulation unless hormonally- varies with time of insemination and those in which it does
induced improvements in penetrability are reduced by the not. The paper then concludes with a discussion of the variables
debris left by earlier inseminations. which control the duration of the follicular phase, showing
Key words: cervical mucus/coital rate/follicular length/sex of that mucus penetrability increases and Y-selection decreases
offspring/time of insemination in longer follicular phases.

Mucus penetrability and sperm selection

Introduction Roberts’ (1978) idea that selection for Y-spermatozoa decreases
Weinberg et al. (1995) report data in which the sex of human as mucus penetrability improves towards the end of the
offspring varies with the length of the follicular phase. They follicular phase has received little acceptance. Among the
report a mean phase length of 15.4 days in 69 menstrual cycles reasons for this are (i) subsequent workers could not replicate
which produced males and a mean of 17.6 days in 64 cycles the finding of Ericsson et al. (1973) that Y spermatozoa
producing females. This statistically significant difference migrate more readily through an albumin gradient (Martin,
raises two interrelated questions. First, what accounts for the 1994) and (ii) the sex of offspring does not always vary with
association between length of the follicular phase and sex of the time of insemination. However, a recent study (Pyrsak,
offspring and, secondly, how do these findings relate to data 1994) is consistent with earlier studies (Ericsson et al., 1973;
which suggest that the sex of offspring varies with (i) the time Rohde et al., 1973; Broer et al., 1976; Beernink et al., 1993;
of insemination relative to ovulation (Guerrero, 1974; Perez Ericsson, 1994) which find that some Y spermatozoa migrate
et al., 1982; France et al., 1984) and with (ii) coital rate more readily through a variety of media, including cervical
(MacMahon and Pugh, 1954; Brewis, 1993; Underwood, mucus, and it answers the criticisms of Ericsson’s work. Pyrsak
1993; Borgerhoff Muldur, 1994; Martin, 1994; Wadley and (1994) reports that, from a population comprising 50% Y
Martin, 1997). spermatozoa, Y spermatozoa make up 66–71% of the first
Weinberg et al. (1995) suggest a number of mechanisms 10% of the spermatozoa which swim through an albumin
which might link the length of the follicular phase to the sex gradient. When more than 10% of the original population
of the baby. One of these is that the properties of cervical passes through the filter, the Y/X ratio approaches that of the
mucus may differ in cycles with longer as against shorter original population. The studies which failed to replicate
follicular phases and that these differences and differences in Ericsson’s results recovered more than the first 10% and their
sperm motilities separate X and Y spermatozoa. Models of failure to find Y selection was therefore to be expected.
sperm selection in which the properties of X and Y spermatozoa Pyrsak’s findings in conjunction with those of Ericsson et al.
interact with the properties of cervical mucus to affect the sex (1973), Rohde et al. (1973), Broer et al. (1976) and Beernink
of offspring have been offered before. Roberts (1978) has et al. (1993) suggest that Y selection should increase with the
suggested that decreases in the viscosity of cervical mucus in severity of filtration by the cervical mucus. It is well known
the late follicular phase and the superior motility of Y (Moghissi, 1973; Odeblad, 1973; Speroff et al., 1989; Wood,
spermatozoa may account for data (Guerrero, 1974) in which 1994) that the penetrability of cervical mucus improves over
the male:female sex ratio at birth declines as the day of the latter stages of the follicular phase, then deteriorates after
responsible insemination approaches ovulation day. ovulation when the corpus luteum begins producing more
© European Society for Human Reproduction and Embryology 611
J.F.Martin

progesterone. Selection for Y spermatozoa should thus be When this method fails because an error or a longer than
highest when the mucus first begins to open up 4 days or so normal follicular phase leads a couple to resume coitus too
before ovulation (Moghissi, 1973; Wood, 1994), decline rapidly soon, the initial coital act in the fertile period will follow a
with the surge in mucus penetrability over the day or two period of abstinence during which time the cervix will have
leading up to ovulation (Templeton et al., 1982), and then rise been cleared of debris and it will occur near the end of the
again after ovulation. fertile period. If the act is pre-ovulatory, endogenous mucus
However, while Weinberg et al. (1995) report an association penetrability will also be good and Y selection will be at the
between sex ratios and length of the phase, they do not find minimum. If the act is post-ovulatory but within the fertile
that the sex ratio at birth declines significantly as the time period, the thickening mucus should result in a higher level
of responsible insemination approaches ovulation. Moreover, of Y selection.
there are other studies in which the sex of offspring does not If, on the other hand, a couple inadvertently continues
vary with time of insemination. Spira et al. (1993) find no intercourse into the beginning of the fertile period, the acts in
association between sex of the baby and time of insemination in the fertile period will follow previous acts which may have
a study of over 1000 births and, contrary to some interpretations left debris in the cervix and will occur while endogenous
(James, 1994, 1995; Weinberg et al., 1995; Jongbloet et al., mucus penetrability is still poor. Aggregating conceptions
1996), Harlap’s (1979) data also fail to show any significant which result from rhythm failures should thus produce a
relation between time of insemination and sex of offspring distribution in which the percentage of male births is quite
(Martin, 1995). high when insemination is early and which then declines as
Martin (1995) argues that the differences between the the time of insemination approaches ovulation before rising
studies in which sex of offspring does correlate with time of after ovulation. The data in these two studies are consistent
insemination (Guerrero, 1974; Perez et al., 1982; France et al., with these expectations. The 875 births in Guerrero’s study
1984) and those in which it does not (Harlap, 1979; Spira (1974) are attributed to inseminations which occurred as early
et al., 1993; Weinberg et al., 1995) are a function of coital as 9 days before ovulation and as many as 3 days after ovulation.
rate. He notes that Baker and Bellis (1993) find that the The percentage of male offspring declines significantly (χ2
penetrability of cervical mucus is affected by more than its for a linear trend 5 17.0, P , 0.001) from day –8 to ovulation.
hormonally-controlled viscosity, vibrosity, water content and After ovulation the percentage of male offspring rises, but not
flow. In addition to its hormonally-induced properties, the significantly. However, it is unlikely that inseminations more
penetrability of mucus is a negative function of the density of than 6 days before or 2 days after ovulation will produce
sperm cells, leukocytes and other debris left in the cervix by conceptions. This and the fact that the subjects in this study
previous inseminations. Over time such debris is flushed from were using the unreliable shift in basal body temperature
the cervix by the flow of fresh mucus. If insemination occurs method to estimate the time of ovulation suggest that these
before the cervix is cleansed, any exogenously-produced conceptions actually resulted from inseminations closer to the
decline in mucus penetrability may be exacerbated. Insemina- time of ovulation. When the births are grouped into those
tion early in the fertile period may thus both raise the attributed to inseminations before the most fertile period (days
probability that any conception will be due to the earlier rather ø–2), the most fertile period (–1, 0) and post-ovulatory
than a later insemination and increase Y selection in later inseminations (ù1), the percentages of male offspring are
inseminations. Increasing the coital rate should therefore 57.3, 45.3 and 50.0. This distribution is still statistically
dampen or eliminate any association between sex of offspring significant (χ2 5 10.45, P , 0.01) and, consistent with
and time of insemination which would otherwise follow expectations, it is U-shaped.
from hormonally-induced improvements in mucus penetrability The smaller (n 5 52) study of Perez et al. (1982) reveals a
(Martin, 1995). similar pattern. The percentage of males in births produced by
The revised or synthetic idea is thus that lower mucus inseminations before day –1 is 70. The percentage is 36.8 for
penetrability selects for Y spermatozoa and that mucus penet- days –1, 0 and rises to 85 for days ù1. There are too few of
rability is controlled by hormonal factors, which increase the post-ovulatory inseminations to evaluate statistically, but
penetrability toward the end of the phase, and by coital rate when the data are aggregated into the groups days –1, 0 and
which, when raised, lowers mucus penetrability. This idea is the rest, the differences are statistically significant (χ2 5 7.65,
consistent with the data on sex of offspring and time of P , 0.01).
insemination and it may also account for the propensity for A similar, but not significant U-shaped regression of the
sex ratios at birth to decline as the follicular phase lengthens percentage of male offspring over time of insemination is
(Weinberg et al., 1995). The rest of the paper is given over to reported by France et al. (1984). In that small study (n 5 33)
documenting these claims. the subjects engaged in coitus only once in the fertile period.
Under these conditions we should expect the percentage of
males to decline and then rise in response to hormonally-
Coital rate, time of insemination and sex of offspring induced changes in mucus penetrability. While there is but one
There are two studies in which the sex of offspring varies birth credited to post-ovulatory inseminations, the percentage of
significantly with the time of insemination. These two studies males does decline from 83.3 for days –5 and –4 to 76.9 for
(Guerrero, 1974; Perez et al., 1982) both involve births which days –3 and –2 and to 53.8 for days –1 and 0. The researchers
result from failures in the rhythm method of birth control. report that a Wilcoxon–Mann–Whitney rank order test produces
612
 Follicular phase and sex of offspring

Table I. Coital pattern, time of insemination and sex of offspring

Study Day of Births Significance Coital pattern

insemination
M F %M

Guerrero, 1974 ø–2 235 175 57.3 χ2 5 10.45 Rhythm failure

–1, 0 139 168 45.3 P ,0.01
ù1 79 79 50.0
Perez et al., 1982 ø–2 14 6 70.0 For –1.0 vs rest Rhythm failure
–1, 0 7 12 36.8 χ2 5 7.65
ù1 11 2 85.0 P ,0.01
France et al., 1984 ø–2 15 4 78.9 For –1, 0 vs rest One insemination in fertile period
–1, 0 7 6 53.8 χ2 5 1.59
ù1 0 1 0.0 P .0.10
Harlap, 1979 ø–2 2728 2423 53.0 χ2 5 0.92 Abstain through menstruation plus at least 7 days,
–1, 0 788 739 51.6 P .0.90 then multiple inseminations
ù1 464 424 52.3
Spira et al., 1993 –6, –5 – – 56.3 Seeking conception, multiple inseminations
–3 – – 38.7
–2 – – 58.5
–1 – – 44.1
0 – – 53.9
11 – – 45.4
12, 13 – – 46.8
Weinberg et al., 1995 ø–2 15 11 57.7 For ø–2, and –1, 0 Seeking conception, multiple inseminations
–1, 0 51 50 50.5 χ2 5 0.43
ù1 – – – P .0.50

a Z score of 1.87, P 5 0.06. The χ2 for days –1 and 0 versus and for at least 7 days thereafter. If they indeed did do this,
the rest is 1.59, P . 0.10. The data from these three studies the first inseminations after the resumption of intercourse were
are summarized in Table I. Table I also includes data from unaffected by debris from previous inseminations. The sex
three studies in which multiple acts in the fertile period should ratio among conceptions from these initial acts should therefore
have dampened any association between time of insemination reflect endogenous mucus penetrability and form a U-shaped
and sex of offspring. In the large (n 5 1052) GEFCO study regression over the days of insemination. Spermatozoa insemin-
(Spira et al., 1993) and the smaller (n 5 127) study of ated subsequently would face debris from the earlier coital
Weinberg et al. (1995) women were trying to conceive. When acts, however, and the sex ratio among conceptions due to
this is the intention, the probability of intercourse increases as subsequent pre-ovulatory inseminations would decline less as
the time of ovulation is approached. In the Weinberg study the time of insemination approached ovulation than it would
women engaged in coitus a mean of 2.5 times in the period day – had there been no previous acts. Debris from earlier pre-
5 through ovulation day. Under these conditions spermatozoa ovulatory inseminations would also exacerbate the strong,
inseminated early in the fertile period face poor endogenous hormonally-induced decline in mucus penetrability which
mucus penetrability and, perhaps, debris from earlier insemina- immediately follows ovulation and reduce the probability of
tions. They also then dampen the improvements in mucus conception from post-ovulatory inseminations. Conceptions
penetrability which normally occur nearer the time of ovulation. which do arise from post-ovulatory inseminations should thus
Y-selection should thus be high for spermatozoa inseminated come largely from coital acts which are not preceded by earlier
early and only marginally less for those introduced later and acts in that fertile period. Sex ratios among conceptions due
there should be little if any systematic variation in the to those late inseminations should therefore be affected by
percentage of males by time of insemination. only endogenous mucus penetrability.
The data from both these studies are consistent with this The births in Harlap’s study are classified according to
expectation. In the Weinberg study the percentage of male when, relative to ovulation, intercourse was resumed. The time
offspring is 57.7 (15 M, 11 F) for days ø–2 and 50.5 for of ovulation is estimated by subtracting 14 days from the mid-
days –1 and 0 (χ2 5 0.43, P . 0.50). Spira et al. (1993) point of the lengths of the last three menstrual cycles. If the
report only sex ratios for days –6, –5 and –3 through ù2. The fertilizing act was always the initial act, the percentage of
values for these days, expressed as percentage of males, are males among Harlap’s 7566 births should form a U-shaped
56.3, 38.7, 58.5, 44.1, 53.9, 45.4, and 46.8. The authors regression over the day of insemination. The percentage of
conclude that there is no relationship between time of insemina- males for cycles classified by the days in which intercourse
tion and sex of offspring. was resumed are 52.8, 53.3, 53.3, 49.3, 50.2, 65.5 and 49.3.
Harlap (1979) has data on 7566 births to women who These are for 3 or more days before ovulation through days
claimed to have abstained from intercourse during menstruation ù 3. The distribution is not U-shaped. However, Harlap
613
J.F.Martin

excludes from the analysis the 430 births (percentage of finding that longer follicular phases are associated with lower
males 5 49.3) from days ù 3 after ovulation and the 3478 secondary sex ratios.
births (52.8% male) from days ø–3. This leaves a distribution
in which the percentage of male offspring declines from 53.3
on days –2 and –1 to 49.3 and 50.2 on days 0 and 11 and Length of the follicular phase and sex of offspring
which then rises to 65.5 on days 12. The variation in this There are a number of theories concerning the variables which
roughly U-shaped distribution is statistically significant and it control the duration of the follicular phase. The first of these
is the distribution which some authors (James 1994, 1995; suggests that the factors underlying variation in the length of
Weinberg et al., 1995; Jongbloet et al., 1996) consider consist- the phase are different concentrations of follicle stimulating
ent with the proposition that regression of the sex ratios at hormone (FSH) in the latter stages of the previous luteal phase
birth over time of insemination is U-shaped. and in the very early follicular phase. While initial follicular
Harlap discards the 430 births from days ù3 because she growth appears to occur independently of FSH, luteinizing
believes the estimated days of insemination to be in error, hormone (LH) and oestradiol, a rise in FSH (and LH) following
probably because inseminations 3 or more days after ovulation the decline in progesterone and oestradiol in the previous
are not likely to produce conceptions. However, this improbab- luteal phase stimulates further follicular development (Speroff
ility also extends to inseminations in cycles in which coitus et al., 1989). A varying number of follicles are stimulated to
was resumed on day 12. The appropriate procedure in these begin development, of which one, perhaps the largest, emerges
cases would be either to discard all these data or, better, pool as the primary follicle. As the primary follicle grows, it
them with the conceptions from cycles in which intercourse produces increasing quantities of oestradiol which, in turn,
was initiated on day 11. Either procedure, however, flattens begin to suppress the production of FSH by the pituitary.
the distribution of sex ratios and eliminates the statistical Because the intrafollicular concentration of oestradiol in the
significance (χ2 5 4.09, P . 0.30 and χ2 5 3.33, P . 0.50, primary follicle is high and because that high concentration
respectively). of oestradiol previously induced more FSH receptors and
Harlap’s reason for also discarding the 3478 births from oestradiol-producing granulosa cells, the primary follicle con-
cycles where intercourse was resumed 3 or more days before tinues to grow at an exponential rate even as its production of
ovulation is that, after several weeks of abstinence, couples oestradiol continues to suppress FSH production by the pituit-
were likely to engage in intercourse on successive days. In ary. The smaller follicles, on the other hand, have fewer FSH
these cases conceptions may have resulted from inseminations receptors, fewer oestradiol-producing granulosa cells, and more
closer to ovulation than day –3. However, this reasoning may androgen-rich intrafollicular environments. Therefore, as FSH
also apply to cycles in which intercourse was resumed on production declines in response to the growth of the primary
days –2 and –1. If, in recognition of these possibilities, days follicle and its increasing production of oestradiol, the smaller
ø–3 and –2 are pooled as are days –1 and 0, the percentages follicles slow their growth and then become atretic (Speroff
male for days ø–2, –1, 0 and ù1 are 53.0, 51.6 and 52.3. et al., 1989; Fowler and Templeton, 1996).
The differences are not statistically significant (χ2 5 0.92, This is significant for present concerns because the small
P . 0.90). follicles produce gonadotrophin surge inhibiting or attenuating
Thus, the differences between those studies in which the factor (GnSIF) (Fowler and Templeton, 1996). GnSIF is
sex of offspring varies significantly with the time of insemina- believed to reduce pituitary responsiveness to increases in the
tion and those in which it does not, seem to rest in the effects pulse frequency and amplitude of gonadotrophin-releasing
of the coital rate just before and during the fertile period. This hormone (GnRH). In this model, increases in GnRH pulse
is evident when the data from Guerrero (1974), Perez et al. frequency normally occur after oestradiol rises continuously
(1982) and France et al. (1984) are aggregated and compared for 2–4 days (WHO, 1980) and/or surpasses a concentration
with the data of Harlap (1979) and Weinberg et al. (1995). of ù200 pg/ml for ~60 h (Speroff et al., 1989; Shoham et al.,
Multiple inseminations just before and early in the fertile 1995). However, for the pituitary to respond to the rise in
period among the first group raise the percentage of male GnRH pulse frequency with the LH surge which reinitiates
offspring for the days ø–2 to 58.8, significantly higher meiosis and is the precursor to ovulation, the effects of GnSIF
(χ2 5 5.61, P 5 0.018) than the 53.0% in the second group, produced by small follicles must be overcome. This occurs as
in which Y selection in the early period is largely a function the oestradiol concentration surges, producing higher GnRH
of endogenous sperm penetrability. At the same time, coitus pulse frequencies, and with a decline in the production of
near the time of ovulation is less frequent in the first group GnSIF as the small follicles atrophy (Shoham et al., 1995;
than it is in the second and the percentage of male offspring Fowler and Templeton, 1996).
for days –1 and 0 is also significantly lower in the first group Accordingly, when higher concentrations of FSH in the
(45.1 versus 51.5; χ2 5 4.60, P 5 0.032). previous luteal phase and the early follicular phase produce
These variations in sex ratios thus seem to result from greater numbers of small follicles, GnSIF concentrations are
variations in mucus penetrability. In the cases at hand, coital higher, take longer to decline, and are associated with and
patterns and hormonally-induced changes over the course of require more sustained and higher amounts of oestradiol to
the follicular phase appear to be the major determinants of induce the LH surge. When the number of FSH-induced small
mucus penetrability. However, the length of the follicular follicles reaches supra-normal concentrations, the LH surge
phase may also affect mucus penetrability and account for the may be inhibited altogether and the cycle will be anovulatory.
614
 Follicular phase and sex of offspring

Conversely, if there are fewer small follicles, earlier and even ovulate. Over the next 7 and 10 days respectively their blood
premature LH surges may be induced. The latter can also was sampled every 4 h. Neither produced an LH surge while
result in anovulation or in an inadequate luteal phase (Sherman hospitalized, but both did so within 48 h of being discharged.
and Korenman, 1974; Messinis and Templeton, 1986; Mroueh The stress associated with hospitalization and blood sampling
et al., 1996). prolonged the follicular phase by at least 7 days in one case
The point here is that, from this interpretation, the number and 10 in the other.
of small follicles may shorten or lengthen the follicular phase In summary, because the LH surge can be advanced or
by advancing or retarding the LH surge. When the LH surge retarded relative to the degree of follicular development and
is advanced ovulation occurs earlier. Correlatively, oestradiol production of more penetrable mucus, mucus penetrability
concentrations will be lower and will not have been sustained should vary with the length of the phase. Shorter phases
for as long. As mucus penetrability parallels the rise in the should be associated with less penetrable mucus and a higher
concentration of oestradiol, shorter follicular phases induced proportion of males among offspring, longer phases with
by fewer small follicles should be associated with a shorter longer periods of more penetrable mucus and a lower proportion
period of mucus penetrability and mucus of poorer quality. of males.
Conversely, if higher concentrations of GnSIF produced by More frequent inseminations, which can reduce or eliminate
larger numbers of small follicles retard but do not attenuate the connection between time of insemination and sex of
the LH surge, the follicular phase will be longer and will be offspring, should not materially affect this relationship.
associated with longer periods of high oestradiol production Increased coital rates will raise Y selection in both longer and
and more penetrable cervical mucus. In summary, if variations shorter phases. For a higher coital rate to reduce the difference
in the length of the follicular phase are due to variations in in sex ratio between longer and shorter phases, it would have
the number of small follicles, shorter phases will be associated to raise the male:female ratio in the longer phase proportionally
with shorter periods of penetrable mucus and higher proportions more than it raised it in the shorter phase. This seems unlikely,
of Y spermatozoa reaching the uterus. Conversely, longer particularly given that the higher rate of flow of cervical mucus
phases will be associated with more penetrable mucus over in the longer phase should cleanse the cervix faster.
longer periods of time and lower proportions of Y spermatozoa
in the uterus.
There is a second, less complex, model of the timing of the
Conclusions
LH surge and ovulation. Levran et al. (1995) found that in 65
women with ovarian failure the LH surge occurred after a Evidence exists that less penetrable mucus selects for Y
cessation of the increase in serum oestradiol. This was achieved spermatozoa and that such selection decreases as mucus
by administering incremental doses of oestradiol valerate for penetrability improves. Endogenous or hormonally-induced
4–8 days. The LH surge did not occur as long as the oestradiol mucus penetrability improves over the late follicular phase
concentration was increasing; LH surges occurred only after and with the length of the follicular phase. However, there is
treatment was suspended and oestradiol concentrations moder- also evidence that debris from previous inseminations reduces
ated. Since these women had no ovarian follicles, it does not mucus penetrability for subsequent inseminations. Higher or
appear that GnSIF could have played any role in the inhibition increasing coital rates should therefore decrease mucus penet-
of the LH surge. This finding, that it is a decline or levelling rability for later inseminations in both shorter and longer
off of oestradiol production which induces the LH surge, has phases. When insemination occurs only once over the 6 or so
the advantage of simplicity and is also consistent with the days before ovulation, or when the coital rate is declining
general propensity for oestradiol to inhibit LH secretion. over that period, hormonally-induced improvements in mucus
However, in both of these models longer follicular phases penetrability should result in male:female ratios which decline
imply longer periods of increasing oestradiol production and as the time of insemination approaches the end of the phase
more penetrable cervical mucus. and as the phase is lengthened. If intercourse is frequent in
Variations in the length of the follicular phase which the period before ovulation, there should be less or no
stem from stress-induced alterations of reproductive functions association between time of insemination and sex of offspring.
produce similar results. It is well known that acute, transient However, while more frequent intercourse should raise the sex
stress may stimulate an LH surge and, derivatively, ovulation. ratio among offspring in both longer and shorter phases, it
Long term or chronic stress, on the other hand, seems to should not alter the propensity for longer phases to be associated
suppress LH production by increasing corticotrophin-releasing with a lower proportion of male offspring because the coital
factor (CRF) and the corticotrophins in turn inhibit the secretion rate affects both longer and shorter phases.
of GnRH by the hypothalamus (Rivier and Rivest, 1991). While the discussion here includes an interactive factor,
Transient stress is thus capable of advancing the LH surge coital rate, it fails to consider any purely male factors such as
while chronic stress can retard it and even result in anovulation. the Y/X ratios in the inseminate and to assess the effects of
Chronic stress may result from poor nutrition, intense physical early spontaneous abortions. Hence, while the factors under
activity, or emotional or psychological factors. As an example consideration may offer some understanding of sources of
of the latter and its effect on the timing of the LH surge, variability in sex ratios, they cannot account for specific levels,
Peyser et al. (1973) report on two regularly ovulating women especially when sex ratios are lower than the Y/X ratio in the
admitted to the hospital just prior to when they were due to inseminate.
615
J.F.Martin

References Sherman, B.M. and Korenman, S.G. (1974) Measurement of plasma LH, FSH,
estradiol and progesterone in disorders of the human menstrual cycle: the
Baker, R.R. and Bellis, M.A. (1993) Human sperm competition: ejaculate short luteal phase. J. Clin. Endocrinol. Metab., 38, 89–93.
manipulation by females and a function for the female orgasm. Animal Shoham, Z., Weissman, A., Schachter, M. et al. (1995) The luteinizing
Behav., 46, 887–909. hormone surge—the final stage in ovulation induction: modern aspects of
Beernink, F.J., Dimowski, W.P. and Ericsson, R.J. (1993) Sex preselection ovulation triggering. Fertil. Steril., 64, 237–251.
through albumin separation. Fertil. Steril., 59, 382–386. Speroff, L., Glass, R.H., and Kase N.G. (1989) Clinical Gynecologic
Borgerhoff Muldur, M. (1994) On polygyny and sex ratio at birth: An Endocrinology and Infertility, 4th edn. Williams and Wilkins, Baltimore.
evaluation of Whiting’s study. Curr. Anthropol., 35, 625–627. Spira, A., Ducot, B., Guihard-Moscato, M. et al. (1993) Conception probability
Brewis, A.A. (1993) Sex ratios at birth in a micronesian atoll population. Soc. and pregnancy outcome in relation to age, cycle regularity, and timing of
Bio., 50, 207–214. intercourse. In Gray, R., Leridon, H. and Spira A. (eds), Biomedical
Broer, K.H., Winkhaus, I., Sombroek, H. and Kaiser, R. (1976) Frequency of and Demographic Determinants of Reproduction. Clarendon, Oxford, pp.
Y-chromatin bearing spermatozoa in intracervical and intrauterine postcoital 271–284.
tests. Internat. J. Fertil., 21, 181–185. Templeton, A.A., Penny, G.C. and Lees, M.M. (1982) Relation between the
Ericsson, R.J. (1994) Sex selection via albumin columns: 20 years of results. luteinizing hormone peak, the nadir of the basal body temperature and the
Hum. Reprod., 9, 1787–1788. cervical mucus score. Br. J. Obstet. Gynecol., 89, 985–988.
Ericsson, R.J., Langevin, C.N. and Nishino, M. (1973) Isolation of fractions Underwood, J. (1993) Secondary sex ratios in micronesian populations. Soc.
rich in human Y-sperm. Nature, 246, 421–424. Bio., 40, 200–206.
Fowler, P.A. and Templeton, A. (1996) The nature and functions of putative Wadley, R. and Martin, J.F. (1997) Secondary sex ratios and coital frequency
gonadotrophin surge-attenuating/inhibiting factor (GnSAF/IF). Endocrine with an Iban case. Curr. Anthropol., 38, 79–81.
Reviews, 17, 103–120. Weinberg, C.R., Baird, D.D. and Wilcox, A.J. (1995) The sex of the baby
France, J.T., Graham, F., Gosling, L. et al. (1984) A prospective study of the may be related to the length of the follicular phase in the conception cycle.
pre-selection of the sex of offspring by timing intercourse relative to Hum. Reprod., 10, 304–307.
ovulation. Fertil. Steril., 41, 894–900. WHO (1980) Temporal relationships between ovulation and defined changes
Guerrero, R. (1974) Association of the type and time of insemination within in the concentration of plasma estradiol-17β, luteinizing hormone, follicle
the menstrual cycle with the human sex ratio at birth. New Eng. J. Med., stimulating hormone, and progesterone. Am. J. Obstet. Gynecol., 138,
291, 1056–1059. 383–390.
Harlap, S. (1979) Gender of infants conceived on different days of the Wood, J.W. (1994) Dynamics of Human Reproduction. Biology, Biometry,
menstrual cycle. New Engl. J. Med., 300, 1445–1448. Demography. Aldine De Gruyter, New York.
James, W.H. (1994) Cycle day of insemination, sex ratio of offspring and
Received on August 9, 1996; accepted on January 7, 1997
duration of gestation. Ann. Hum. Bio., 21, 263–266.
James, W.H. (1995) Follicular phase length, cycle day of conception and sex
ratio of offspring. Hum. Reprod., 10, 2529–2531.
Jongbloet, P.H., Groenewoud, J.M.M. and Zielhuis, G.A. (1996) Non-optimal
maturation of oocytes and the sex ratio. Hum. Reprod., 11, 2–9.
Levran, D., Ben-Shlomo, I., Luski, A. et al. (1995) A reappraisal of the
feedback effects of oestradiol upon luteinizing hormone surge. Hum.
Reprod., 10, 3117–3120.
MacMahon, B. and Pugh, T.F. (1954) Sex ratio of white births in the US
during the second world war. Am. J. Hum. Genet., 6, 284–292.
Martin, J.F. (1994) Changing sex ratios. The history of Havasupai fertility
and its implications for human sex ratio variation. Curr. Anthropol., 35,
235–279.
Martin, J.F. (1995) Hormonal and behavioral determinants of the secondary
sex ratio. Soc. Bio., 42, 226–238.
Martin, R. (1994) Human sex pre-selection by sperm manipulation. Hum.
Reprod., 9, 1790–1791.
Messinis, I.E. and Templeton, A. (1986) The effect of pulsatile follicle
stimulating hormone on endogenous luteinizing hormone surge in women.
Clin. Endocrinol., 35, 403–407.
Moghissi, K.S. (1973) Sperm migration through the human cervix. In Elstein,
M., Moghissi, K.S. and Borth, R. (eds), Cervical Mucus in Human
Reproduction. Scriptor, Copenhagen, pp. 128-152.
Mroueh, J.M., Arbogast, L.K., Fowler, P. et al. (1996) Identification of
gonadotrophin surge-inhibiting factor (GnSIF)/attenuin in human follicular
fluid. Hum. Reprod., 11, 490–496.
Odeblad, E. (1973) Biophysical techniques of assessing cervical mucus and
microstructure of cervical epithelium. In Elstein, M., Moghissi, K.S.
and Borth, R. (eds), Cervical Mucus in Human Reproduction. Scriptor,
Copenhagen, pp. 58–74.
Perez, A., Eger, R., Domenichini, V. et al. (1982) Sex ratio associated with
natural family planning. Fertil. Steril., 43, 152-153.
Peyser, M.R., Avalon, D., Harell, A. et al. (1973) Stress induced delay of
ovulation. Obstet. Gyn., 42, 667–671.
Pyrsak, R. (1994) Separation of X- and Y-bearing spermatozoa using albumin
gradients. Hum. Reprod., 9, 1788–1790.
Rivier, C. and Rivest, S. (1991) Review. Effect of stress on the activity of the
hypothalmic–pituitary–gonadal axis: peripheral and central mechanisms.
Bio. Reprod., 45, 523–532.
Roberts, A.M. (1978) The origins of fluctuations in the human secondary sex
ratio. J. Biosoc. Sci., 10, 169–182.
Rohde, W., Porstmann, T. and Dorner, G. (1973) Migration of Y-bearing
spermatozoa in cervical mucus, J. Reprod. Fertil., 33, 167–169.

616