Fraternal birth order and the maternal immune hypothesis of male homosexuality

In men, sexual orientation correlates with an individual's number of older brothers, each additional older brother increasing the odds of homosexuality by approximately 33%. It has been hypothesized that this fraternal birth order effect reflects the progressive immunization of some mothers to Y-linked minor histocompatibility antigens (H-Y antigens) by each succeeding male fetus and the concomitantly increasing effects of such maternal immunization on the future sexual orientation of each succeeding male fetus. According to this hypothesis, anti-H-Y antibodies produced by the mother pass through the placental barrier to the fetus and affect aspects of sexual differentiation in the fetal brain. This explanation is consistent with a variety of evidence, including the apparent irrelevance of older sisters to the sexual orientation of later born males, the probable involvement of H-Y antigen in the development of sex-typical traits, and the detrimental effects of immunization of female mice to H-Y antigen on the reproductive performance of subsequent male offspring. The maternal immune hypothesis might also explain the recent finding that heterosexual males with older brothers weigh less at birth than heterosexual males with older sisters and homosexual males with older brothers weigh even less than heterosexual males with older brothers.     

Fraternal birth order and birth weight in probably prehomosexual feminine boys

The purpose of this study was to confirm a previous finding that homosexual males with older brothers weigh less at birth than do heterosexual males with older brothers. The subjects comprised 250 feminine boys referred to a child psychiatry service because of extreme cross-gender wishes or behavior and assumed, on the basis of previous research, to be prehomosexual, plus 739 control boys and 261 control girls referred to the same service for reasons unrelated to sexual orientation or gender identity disorder and assumed, from base-rate probabilities, to be preheterosexual. The feminine boys with two or more older brothers weighed 385 g less at birth than did the control boys with two or more older brothers (P = 0.005). In contrast, the feminine and control boys with fewer than two older brothers did not differ in birth weight. This finding suggests that the mechanism by which older brothers increase the odds of homosexuality in later-born males operates prior to the individual's birth. We hypothesize that this mechanism may be immunologic, that antimale antibodies produced by human mothers in response to immunization by male fetuses could decrease the birth weight of subsequent male fetuses as well as increase their odds of homosexuality.     

Quantitative and theoretical analyses of the relation between older brothers and homosexuality in men

Meta-analysis of aggregate data from 14 samples representing 10,143 male subjects shows that homosexuality in human males is predicted by higher numbers of older brothers, but not by higher numbers of older sisters, younger brothers, or younger sisters. The relation between number of older brothers and sexual orientation holds only for males. This phenomenon has therefore been called the fraternal birth order effect. Research on birth order, birth weight, and sexual orientation suggests that the developmental pathway to homosexuality initiated by older brothers operates during prenatal life. Calculations assuming a causal relation between older brothers and sexual orientation have estimated the proportion of homosexual men who owe their sexual orientation to fraternal birth order at 15% in one study and 29% in another. The maternal immune hypothesis proposes that the fraternal birth order effect reflects the progressive immunization of some mothers to male-specific antigens by each succeeding male fetus and the increasing effects of such immunization on sexual differentiation of the brain in each succeeding male fetus. There are at least three possible mechanisms by which the mother's immune response could influence the fetus: the transfer of anti-male antibodies across the placenta from the maternal into the fetal compartment, the transfer of maternal cytokines across the placenta, and maternal immune reactions affecting the placenta itself. This hypothesis is consistent with recent studies showing that the quantity of fetal cells that enter the maternal circulation is greater than previously thought, and that the number of male-specific proteins encoded by Y-chromosome genes is greater than previously thought.     

The influence of birth order on birth weight: does the sex of preceding siblings matter?

The purpose of this study was to investigate whether the well-established relationship between parity and birth weight is affected by the sex composition of siblings, especially for male newborns. Subjects were 856 male and 862 female newborns who weighed at least 2500 g at birth, who were born after 37 completed weeks of gestation, who obtained an Apgar score of 7 or higher, who had the same biological parents as all other children in the sibship, and who lived in the same household. Information on birth weight was collected from hospital records. Results showed that male newborns with older brothers weighed less than male newborns with older sisters. In contrast, the weight of female newborns with older brothers did not differ from the weight of female newborns with older sisters. One explanation of these results is that maternal immunoreactivity to some male-specific feature of the fetus affects prenatal development and consequently reduces birth weight in males. The relation between older brothers and birth weight may have theoretical significance for behavioural variables.     

Evidence for maternally inherited factors favouring male homosexuality and promoting female fecundity

The Darwinian paradox of male homosexuality in humans is examined, i.e. if male homosexuality has a genetic component and homosexuals reproduce less than heterosexuals, then why is this trait maintained in the population? In a sample of 98 homosexual and 100 heterosexual men and their relatives (a total of over 4600 individuals), we found that female maternal relatives of homosexuals have higher fecundity than female maternal relatives of heterosexuals and that this difference is not found in female paternal relatives. The study confirms previous reports, in particular that homosexuals have more maternal than paternal male homosexual relatives, that homosexual males are more often later-born than first-born and that they have more older brothers than older sisters. We discuss the findings and their implications for current research on male homosexuality.     

Sexual orientation, handedness, sex ratio and fetomaternal tolerance-rejection

Fraternal birth order (FBO) appears as a prenatal cause of 15% of homosexual males (gays) through mnemonic maternal anti-male factors. Non-right-handed men seem to be protected from homosexuality. Four hypotheses are proposed: (1) androgenic factors of non-right-handedness neutralize anti-male factors; (2) non-right-handedness and homosexuality are lethal or produce mental impairment; (3) non-right-handed male embryos are insensitive to anti-male factors; (4) mothers of non-right-handed fetuses do not produce anti-male factors. Studies of the sex ratio (SR) of older and younger siblings show: (1) a significant heterogeneity in the SR of siblings of right or non-right handed heterosexual men and women; (2) lesbians are born among siblings with high SR; (3) siblings of right-handed gays show a higher SR than non-right-handed gays that present a low SR. Based on our discovery of maternal tolerance-rejection processes, associated with genetic systems (ABO, Rh), where zygotes or embryos different from their mother induce better pregnancy and maternal tolerance than do those that share antigens with their mothers, I propose a new explanation for sexual relationships, sexual orientation, handedness and sibling SR. Lesbian embryos could induce tolerance from mothers with anti-female factors. Non-right-handedness could induce maternal tolerance, or change the maternal compatibility of "gay" embryos. Alternatively, gay embryos could be poor inducers of maternal tolerance towards male traits.     

The association between the fraternal birth order effect in male homosexuality and other markers of human sexual orientation

Later fraternal birth order (FBO) is a well-established correlate of homosexuality in human males and may implicate a maternal immunization response in the feminization of male sexuality. This has led to the suggestion that FBO may relate to other markers of male sexual orientation which are robustly sexually dimorphic. If so, among homosexual males the number of older brothers should strongly correlate with traits such as spatial ability and psychological gender, indicative of greater behavioural feminization, compared to heterosexual males. The present study failed to find significant associations between number of older brothers and these traits.     

Patrilineal family values,family planning and variation in stature among Taiwanese six-year-olds

It has been argued that patrilineal joint family systems tend to bias family planning decisions in favour of sons. A simple model suggests that in such societies, any given son will be more highly valued by his parents (1) the fewer his brothers and (2) the earlier his birth is in the brother series. A daughter's value will be greater (1) the fewer brothers she has and (2) the earlier her birth is relative to other sisters. This study first addresses the extent of son preference as inferred from family composition data for 772 Taiwanese first-graders born in the mid-1970s in two socioeconomically distinct communities in Taipei, Taiwan. It then uses linear regression to consider whether the model criteria help account for statural variation among children in each study area when controlling for differences in measurement age, parental education and housing. With respect to family composition and gender preference, available evidence was consistent with previous surveys. While better-educated parents in the more affluent study area had significantly fewer children (p < 0.0005) and were more willing to stop without a son, girls there, as in the less affluent area, were still significantly more likely than boys to belong to large sibships (p < or = 0.005). Evidence from mean height of males and females partially accords with hypothetical predictions. In the less affluent area, the interaction effect of male birth order and the presence of younger siblings was significantly associated with mean stature (p = 0.002). Males without brothers were 2.0 cm taller than males with either an older or a younger brother (116.3 +/- 0.5 cm vs 114.3 +/- 0.4 cm). Males who had both younger and older brothers, but often no sisters, were about as tall, however, as those without brothers. A similar, but less pronounced, pattern was found among males in the more affluent area, but only among those who had sisters. These boys were also consistently shorter than boys without sisters (115.6 +/- 0.6 cm vs 117.7 +/- 0.6 cm; p = 0.001). Patterns of mean female stature did not clearly support the hypothesis. Girls in the more affluent area were relatively tall and did not show significant variation. Results among less affluent girls showed significant contrasts, but not necessarily in the predicted direction.     

Testing Miller's theory of alleles preventing androgenization as an evolutionary explanation for the genetic predisposition for male homosexuality

The genetic background of male homosexuality presents an evolutionary paradox, since homosexuality could be considered a reproductive disadvantage. We tested E.M. Miller's (2000) balanced polymorphism explanation, which states that alleles partially preventing androgenization in male fetuses during pregnancy would be associated with a homosexual orientation. Having all the alleles produces homosexuality, while heterosexual carriers with only a few of these alleles instead have a reproductive advantage; that is, they have more traits, which, by controlling for excessive aggressiveness and psychopathy, make them more attractive mates. Pairs of brothers were used to test these assumptions. If homosexuality is due to having all the androgenization-preventing alleles, then heterosexual men with homosexual brothers are more likely to also have some of the these alleles compared to heterosexual men with heterosexual brothers. These two categories were compared on variables related both directly and indirectly to reproductive success, with heterosexual men with a homosexual brother hypothesized to have an advantage on the variables. However, no statistically significant findings in support of the theory were detected. The results were discussed together with alternative explanations.     

Fetal sex determination in infants in Punjab,India: correlations and implications.

OBJECTIVES--To determine the proportion of children whose sex was determined prenatally among those attending one Indian hospital and to identify factors which affect use of fetal sex determination. DESIGN--Cross sectional study using interviews with mothers. SETTING--Medical school hospital in Punjab, India. SUBJECTS--596 children delivered or seen for inpatient or outpatient care. MAIN OUTCOME MEASURES--Fetal sex determination, sex of child, number and sex of siblings, type of care received, socioeconomic status, and maternal education. RESULTS--Sex had been determined prenatally for fewer girls (5/236, 2%) than boys (49/360, 14%). Fetal sex determination had been done for only 2% (3/154) of first born boys compared with 18% (12/66) with one older sister and no older brother and 63% (30/48) with more than one older sister and no older brother. Only four boys whose sex had been determined prenatally had older brothers. The five girls whose sex had been determined prenatally either had a male twin or were incorrectly identified as male. Prenatal sex determination had been done for 21% (26/122) of boys admitted for inpatient care compared with 11% (19/173) seen as outpatients. Use of fetal sex determination increased with increasing monthly income (chi2 for trend = 6.384, P = 0.0115). None of the mothers who had had no education had used fetal sex determination, but among mothers with some education the frequency of use did not change with increasing education. The sex ratio of children born at the hospital rose from 107 boys/100 girls in 1982 to 132 boys/100 girls in 1993. CONCLUSIONS--Fetal sex determination was common, especially if the family already had daughters. Sex determination seems to be driven by a desire to have sons, with socioeconomic status and education having little effect. The lower prevalence of fetal sex determinations for girls is likely to be due to abortion of fetuses found to be female.     

Weight growth in infants born to mothers who smoked during pregnancy.

OBJECTIVE--To determine whether maternal smoking during pregnancy causes impairment in growth after birth. DESIGN--Longitudinal study. SETTING--Six medical university centres of six towns of north, central, and south Italy. SUBJECTS--12,987 babies (10,238 born from non-smoking mothers, 2276 from mothers smoking one to nine cigarettes a day, and 473 from mothers smoking > or = 10 cigarettes a day) entered the study. MAIN OUTCOME MEASURES--Difference in weight gain between children born to smoking mothers and those born to non-smoking mothers. Weight was measured at birth and at 3 and 6 months of age. Maternal smoking habit was derived from interview on third or fourth day after delivery. RESULTS--Compared with children born to mothers who did not smoke during pregnancy, the birth weights of children born to mothers who smoked up to nine cigarettes a day were 88 g (girls) and 107 g (boys) lower; in children born to mothers who smoked > or = 10 cigarettes a day weights were 168 g and 247 g lower. At six months of age for the first group the mean weight for girls was 9 g (95% confidence interval -47 g to 65 g) higher and for boys 64 g (-118 g to -10 g) lower than that of children born to mothers who did not smoke. The corresponding figures for the second group were 28 g (-141 g to 85 g) lower for girls and 24 g (-136 g to 88 g) lower for boys. CONCLUSIONS--The deficits of weight at birth in children born to mothers who smoked during pregnancy are overcome by 6 months of age. These deficits are probably not permanent when smoking habit during pregnancy is not associated with other unfavourable variables (such as lower socioeconomic class).     

The cause(s) of the fraternal birth order effect in male homosexuality

It has been established that the probability that a man is homosexual is positively related to his number of older brothers, but not older sisters when the brothers are accounted for. This is known as the 'fraternal birth order' effect. In the past, efforts have been made to explain this phenomenon in terms of several alternative biological hypotheses and a psychosocial hypothesis. This note examines how well these hypotheses accommodate the fraternal birth order effect. It is concluded that: (1) the evidence for the hypothesis of maternal immunoreactivity to the male fetus is weak; (2) the evidence for the intrauterine hormone exposure hypothesis is also weak; (3) the evidence for the hypothesis of postnatal learning is stronger. Lastly, there seem likely to be causes common to male homosexuality and paedophilia. They may include sexual (or quasi-sexual) experience in childhood or adolescence.     

2D:4D finger-length ratios in children and adults with gender identity disorder

Previous research suggests that prenatal testosterone affects the 2D:4D finger ratio in humans, and it has been speculated that prenatal testosterone also affects gender identity differentiation. If both things are true, then one would expect to find an association between the 2D:4D ratio and gender identity. We measured 2D:4D in two samples of patients with gender identity disorder (GID). In Study 1, we compared the 2D:4D ratios of 96 adult male and 51 female patients with GID to that of 90 heterosexual male and 112 heterosexual female controls. In Study 2, we compared the 2D:4D ratios of 67 boys and 34 girls with GID to that of 74 control boys and 72 control girls. In the sample of adults with GID, we classified their sexual orientation as either homosexual or non-homosexual (in relation to their birth sex) to examine whether or not there were any within-group differences as a function of sexual orientation. In the sample of adult men with GID (both homosexual and non-homosexual) and children with GID, we found no evidence of an altered 2D:4D ratio relative to same-sex controls. However, women with GID had a significantly more masculinized ratio compared to the control women. This last finding was consistent with the prediction that a variance in prenatal hormone exposure contributes to a departure from a sex-typical gender identity in women.     

Sex differences in body composition early in life

Background: Early development of the percentage of fat and muscle is rarely considered, but is important because excessive fat is related to the development of diabetes and other morbidities later in life. In pediatric medicine, there are few to no data comparing sex differences in body composition in the first months of life despite the fact that males are typically longer and weigh more than girls at birth.Objective: The purpose of this study was to determine whether observed sex differences in body composition at birth persist through the first 6 months of life.Methods: Participants were healthy, full-term, male and female newborns. Children throughout the Oklahoma City, Oklahoma, metropolitan area were enrolled. The inclusion criteria were: mothers aged 18 to 45 years at the time of delivery; a term pregnancy lasting ≥37 weeks of gestation (determined by mother's physician); weight adequate for gestational age; and a hospital stay for the infant of <3 days following delivery. The exclusion criteria were: maternal tobacco use or alcohol consumption (>1 drink per week) during pregnancy; gestational diabetes; preeclampsia; and infants with presumed or known congenital birth defects. Baseline assessment at birth included length and weight. Newborns had their body composition (percent fat [%fat], total fat, and fat-free mass) determined at ~1 month of age using whole body plethysmography. Mothers were invited to have their children take part in a 5-month extension that conducted additional body composition measurements at 3 and 6 months of age.Results: Sixty-four girls (mean [SD] age at time of testing, 20.9 [7.9] days; birth weight, 3500 [388] g; birth length, 49.9 [2.4] cm; white race, 73.4%) and 53 boys (mean age at time of testing, 20.2 [7.3] days; birth weight, 3353 [413] g; birth length, 51.0 [2.4] cm; white race, 69.8%) were assessed and included in the study. At birth, girls were significantly shorter and weighed more than boys (both, P < 0.05). At ~1 month of age, body composition revealed that girls had significantly greater %fat (15.1% vs 12.7%; P < 0.05) and less fat-free mass (3182 [303] vs 3454 [361] g; P < 0.001) than did boys. At 3 months of age, girls continued to have significantly less fat-free mass (4379 [347] vs 4787 [310] g; P < 0.01) than did boys; however, by 6 months of age, no significant sex difference was observed in any body composition variable studied.Conclusion: In this small sample of healthy, full-term newborns, at ~1 month of age, statistically significant differences in %fat and fat-free mass existed between girls and boys; however, by 6 months of age, these differences no longer existed.     

Homosexual women have less grey matter in perirhinal cortex than heterosexual women

Is sexual orientation associated with structural differences in the brain? To address this question, 80 homosexual and heterosexual men and women (16 homosexual men and 15 homosexual women) underwent structural MRI. We used voxel-based morphometry to test for differences in grey matter concentration associated with gender and sexual orientation. Compared with heterosexual women, homosexual women displayed less grey matter bilaterally in the temporo-basal cortex, ventral cerebellum, and left ventral premotor cortex. The relative decrease in grey matter was most prominent in the left perirhinal cortex. The left perirhinal area also showed less grey matter in heterosexual men than in heterosexual women. Thus, in homosexual women, the perirhinal cortex grey matter displayed a more male-like structural pattern. This is in accordance with previous research that revealed signs of sex-atypical prenatal androgenization in homosexual women, but not in homosexual men. The relevance of the perirhinal area for high order multimodal (olfactory and visual) object, social, and sexual processing is discussed.     

Birth order and male androphilia in Samoan fa'afafine

The manner in which male androphilia is publicly expressed varies cross-culturally. As such, it is unclear whether distinct or common underlying causal processes characterize male androphilia in different cultures. Establishing the existence of cross-cultural universals in male androphilia is one means of ascertaining whether common biological bases underlie this phenomenon despite its culturally distinct forms. The evidence that the number of older biological brothers increases the odds of androphilia in later-born males has been well documented for Western samples (i.e. the fraternal birth order effect); but there is little evidence for this effect in non-Western samples. Here, we compare the birth order of androphilic males (i.e. fa'afafine) and gynephilic males from the politically autonomous Polynesian nation of Independent Samoa. Results indicate that relative to gynephilic males, fa'afafine tend to have more siblings and are generally later born when birth order is quantified using Slater, fraternal and sororal indices. More specifically, fa'afafine tend to have a greater number of older brothers, older sisters and younger brothers. We discuss the observed effects in relation to the differing reproductive patterns exhibited by the mothers of fa'afafine and gynephilic males, and to existing social and biological theories for sexual orientation.     

Heights and weights of Da-an boys: did sisters really make a difference?

This study further examined the negative association between boys' growth and the presence of sisters within a relatively affluent community in Taipei, Taiwan. Among 596 boys born in 1976-77, differences in height and log-transformed weight were judged using analysis of covariance (ANCOVA) with measurement age as a covariate, and parental education level (four levels), number of sisters (0, 1, 2 or 3+) and number of brothers (0, 1, 2+) as predictors. The relative importance of birth order and sibling sex was examined among the near majority of boys with one sibling (47%, 278/596). The sibling composition variable was defined using mutually exclusive categories representing individuals with one sibling: either one older or younger brother or sister. All boys contributed information before leaving middle school at a mean age of 14.9 +/- 0.4 years SD. The results were compared with similar analyses of data for 154 of these same boys for whom measurements were available from primary school entry at a mean age of 64 +/- 0.3 years SD onward. Results were also compared with data for a cohort of 153 boys who entered primary school later in 1986. Results confirm that boys from the Da-an area born in the mid-1970s who did not have sisters were significantly taller (2.2-2.5 cm, p < or = 0.008) and heavier (3.0-3.9 kg, p < or = 0.016) than those with one or two sisters. However, the 26 boys with three or more sisters, most of whom were last-born, were somewhat taller than those with one or two sisters. The same curvilinear relationships in height and weight appeared both among boys as they prepared to leave middle school and among the subset also measured just after entering primary school. When numbers of sisters were statistically controlled, the presence of two or more brothers was also significantly negatively associated with mean stature, but not weight, among middle school boys. Analyses among boys with one sibling revealed that birth order was associated with mean stature, but only if the sibling was female; an older sister was associated with a greater deficit in mean stature than a younger sister. Evidence of rising educational expectations, continued declines in family size with fewer gender-related differences in numbers of siblings, and a clear secular increase in body size in this community among children entering primary school from 1982 to 1986 suggest a possible explanatory model.     

Mothers' birth weight and survival of their offspring: population based study.

OBJECTIVE: To test the hypothesis that a baby''s survival is related to the mother''s birth weight. DESIGN: Population based dataset for two generations. SETTING: Population registry in Norway. SUBJECTS: All birth records for women born in Norway since 1967 were linked to births during 1981-94, thereby forming 105104 mother-offspring units. MAIN OUTCOME MEASURES: Perinatal mortality specific for weight for offspring in groups of maternal birth weight (with 500 g categories in both). RESULTS: A mother''s birth weight was strongly associated with the weight of her baby. Maternal birth weight was associated with perinatal survival of her baby only for mothers with birth weights under 2000 g. These mothers were more likely to lose a baby in the perinatal period (odds ratio 2.3, 95% confidence interval 1.4 to 3.7). Among mothers with a birth weight over 2000 g there was no overall association between mother''s weight and infant survival. There was, however, a strong interaction between mother''s birth weight, infant birth weight, and infant survival. Mortality among small babies was much higher for those whose mothers had been large at birth. For example, babies weighing 2500-2999 g had a threefold higher mortality if their mother''s birth weight had been high (> or = 4000 g) than if the mother had been small (2500-2999 g). CONCLUSION: Mothers who weighed less than 2000 g at birth have a higher risk of losing their own babies. For mothers who weighed > or = 2000 g their birth weight provides a benchmark for judging the growth of their offspring. Babies who are small relative to their mother''s birth weight are at increased risk of mortality.     

Patterns of low birth weight in the Bengali newborns

This paper examines the distribution of low birth weight (2500 g or less) by gestation time, sex, maternal age, parity (birth order), socioeconomic conditions, and season of birth among 5117 single live births born to Bengali mothers at the Ramakrishna Mission Seva Pratisthan Hospital in Calcutta, India. Preterm infants have low birth weight significantly more often than their full term counterparts. Female infants have low birth weights significantly more often than male infants. The infants of poor mothers have lower birth weights in higher order births more often than infants of higher orders born to well-off mothers. Teenaged mothers produce low birth weight babies significantly more often than older mothers. Although the relationship is not significant, low birth weight infants occur more often among 1st and late born infants and less often among 2nd born infants. The season of birth is not significantly associated with birth weight. Less than 10% of low birth weight infants are pre-term, while the rest are full term. The great majority of low birth weight infants are small-for-gestational-age; the minority are small due to curtailed gestational age. The proportion of infants weighing less than 2001 g is only 9%; this figure tallies closely with earlier studies of India.     

Fecundity of Paternal and Maternal Non-Parental Female Relatives of Homosexual and Heterosexual Men

A variety of social, developmental, biological and genetic factors influence sexual orientation in males. Thus, several hypotheses have attempted to explain the sustenance of genetic factors that influence male homosexuality, despite decreased fecundity within the homosexuals. Kin selection, the existence of maternal effects and two forms of balancing selection, sexually antagonistic selection and overdominance, have been proposed as compensatory mechanisms for reduced homosexual fecundity. Here, we suggest that the empirical support for kin selection and maternal effects cannot account for the low universal frequency and stability of the distribution of homosexuals. To identify the responsible compensatory mechanism, we analyzed fecundity in 2,100 European female relatives, i.e., aunts and grandmothers, of either homosexual or heterosexual probands who were matched in terms of age, culture and sampling strategy. Female relatives were chosen to avoid the sampling bias of the fraternal birth order effect, which occurs when indirectly sampling mothers though their homosexual sons. We observed that the maternal aunts and grandmothers of homosexual probands were significantly more fecund compared with the maternal aunts and maternal grandmothers of the heterosexual probands. No difference in fecundity was observed in the paternal female lines (grandmothers or aunts) from either of the two proband groups. Moreover, due to the selective increase in maternal female fecundity, the total female fecundity was significantly higher in homosexual than heterosexual probands, thus compensating for the reduced fecundity of homosexuals. Altogether, these data support an X-linked multi-locus sexually antagonistic hypothesis rather than an autosomal multi-locus overdominance hypothesis.