- Top of page
- MATERIAL AND METHODS
- LITERATURE CITED
Historically, medical concerns about the deleterious effects of closely inbred marriages have focused on the risk posed by recessive Mendelian disease, with much less attention to developmental instability. We studied the effects of inbreeding (first-cousin marriage) on growth and fluctuating asymmetry of 200 full-term infants (101 inbred and 99 outbred) whose parents were of similar socioeconomic status in Sivas Province, Turkey. In addition to differences in their mean inbreeding coefficients (f = 1/16 for first cousins and f < 1/1,024 for unrelated parents), the consanguineous parents were less well educated (3 years, on average for both husbands and wives). We measured weight, height, head circumference, and chest circumference of the newborns, as well as four bilateral traits (ear width, ear length, and second and fourth digit lengths). After taking education into account, none of the measures of size (weight, height, head circumference, and chest circumference) and fluctuating asymmetry differed between the inbred and outbred groups. Male children of well-educated parents, however, were larger and had less fluctuating asymmetry. Female children of well-educated parents weighed more than those of less well-educated parents, but were otherwise indistinguishable for height, head circumference, chest circumference, and fluctuating asymmetry. We conclude that inbreeding depression causes neither an increase in fluctuating asymmetry of full-term newborns, nor a decrease in body size. Unmeasured variables correlated with education appear to have an effect on fluctuating asymmetry and size of male children and only a weak effect on size (weight) of female children. Am J Phys Anthropol 153:45–51, 2014. © 2013 Wiley Periodicals, Inc.
Inbreeding, the mating of close relatives, reduces the survival and reproductive fitness of normally outbred plants and animals, including humans (Crow and Kimura, 1970; Lynch and Walsh, 1998; Charlesworth and Willis, 2009). Inbreeding exposes deleterious recessive alleles and reduces heterozygosity. The inbreeding depression that results from such consanguineous mating is mostly a consequence of deleterious recessive alleles in the homozygous state (Charlesworth and Willis, 2009). Nevertheless, one cannot discount true overdominance (heterozygote superiority) at a few critical loci, such as those in the major histocompatibility complex (Kekäläinen et al., 2009; Lie et al., 2009; Oliver et al., 2009). Despite the huge literature on inbreeding depression, it is still unclear whether or not inbreeding has a consistent impact on fluctuating asymmetry, a measure of the stability of development under constant environmental and genetic conditions (Zakharov, 1992). Several articles have shown that inbred lines of various animals have greater fluctuating asymmetry than corresponding outbred lines (Mather, 1953; Beardmore, 1960; van Noordwijk and Scharloo, 1981; Leamy, 1984; Alados et al., 1995; Gomendio et al., 2000; Carter et al., 2009). Other studies, though, have shown no effect of inbreeding on fluctuating asymmetry (Keller and Passera, 1993; Fowler and Whitlock, 1994; Gilligan et al., 2000; Hosken et al., 2000; Leamy et al., 2001; Kruuk et al., 2003). In studies of humans, some small, inbred populations have high fluctuating asymmetry (Bailit et al., 1970; Hershkovitz et al., 1993; Markow and Martin, 1993; Schaefer et al., 2006). In addition, Özener (2010) has shown that a group of Turkish high school students whose parents are first cousins weigh less than, and have greater fluctuating asymmetry than, a matched group whose parents are not closely related. Here, we show that inbred human infants of first-cousin marriages do not have greater fluctuating asymmetry than those of outbred ones. Moreover, their size (weight, height, head circumference, and chest circumference) is not reduced. In addition, the parents of the inbred children are less well educated than the parents of the outbred children, which almost certainly confounds fluctuating asymmetry and size with behaviors associated with education. Male children of less well educated parents were smaller and more asymmetrical at birth than those of more educated parents.
Fluctuating asymmetry is the extent to which the average individual departs from perfect symmetry, be it bilateral, radial, or translatory symmetry. It is an indicator of developmental instability and random developmental noise. A large literature suggests that various forms of genetic and environmental stress can increase fluctuating asymmetry, though the patterns are inconsistent (Palmer and Strobeck, 1986; Markow, 1994; Møller and Swaddle, 1997; van Dongen, 2006; Graham et al., 2010). In addition, the genetic basis of fluctuating asymmetry is unclear (Leamy and Klingenberg, 2005), but the resilience of gene regulatory, metabolic, and protein-protein interaction networks operating in noisy developmental systems appears to play a role (Graham et al., 2012; Levy and Siegal, 2008).
- Top of page
- MATERIAL AND METHODS
- LITERATURE CITED
With the exception of the correlations of ear length and width asymmetries with second digit asymmetry (r ≤ 0.091, P ≥ 0.199), the unsigned asymmetries of the four bilateral traits were positively correlated across the entire data set (r ≥ 0.173, P ≤ 0.015). Moreover, Kendall's Coefficient of Concordance (W) was 0.089 (X2 = 53.174, df = 3, P < 0.001) for the entire data set. The same correlations in the four marriage type by sex subpopulations were much less consistent. Nevertheless, three of the four subgroups (inbred and outbred females and inbred males) had significant values of W (W ≥ 0.090, df = 3, P ≤ 0.003) so there was evidence for an individual asymmetry parameter in these groups. In addition, the level of integration among the four bilateral traits was low, even for ear length and width and digits 2 and 4. Among the 24 correlations among these four variables (signed asymmetry), subdivided into four combinations of marriage type and sex, only four correlations were statistically significant (P < 0.05). Of these four correlations, one was negative, the other three were moderately positive. Moreover, the significant correlations were inconsistent with respect to marriage type and sex. Consequently these four traits are reasonably independent.
Descriptive statistics for inbred and outbred newborns are presented in Tables 1 (for males) and 2 (for females). Both male and female parents of outbred infants had, on average, 3.1 years more education than those of the inbred infants. Male parents had 1.36 years more education than female parents. The differences in education (for both fathers and mothers) between marriage types were highly significant (MANOVA, Wilk's λ = 0.815, F2,197 = 22.39, P < 0.001). Moreover, the number of years of education of male and female parents were highly correlated for both inbred (rpearson = 0.692, n = 101, P < 0.001) and outbred (rpearson = 0.816, n = 99, P < 0.001) marriages. After taking total years of education into account (ANCOVA, F1,197 = 30.30, P < 0.001), the total number of pregnancies did not differ between consanguineous and nonconsanguineous couples (ANCOVA, F1,197 = 0.359, P > 0.540).
Table 1. Descriptive statistics (means and standard deviations) for inbred and outbred males
| ||Inbred males||Outbred males|
|Mother's ed. level (year)||5.98||2.53||9.04||3.69|
|Father's ed. level (year)||7.35||3.21||10.45||3.93|
|Head circumference (cm)a||35.49||0.99||35.15||0.97|
|Chest circumference (cm)a||33.82||1.02||33.41||1.20|
|Ear length FA (mm)a||1.31||1.50||1.16||1.12|
|Ear width FA (mm)a||1.26||1.04||0.95||1.06|
|Second digit FA (mm)a||0.53||0.53||0.61||0.68|
|Fourth digit FA (mm)a||0.57||0.65||0.56||0.48|
|Composite FA (mm)a||0.92||0.50||0.82||0.56|
Table 2. Descriptive statistics (means and standard deviations) for inbred and outbred females
| ||Inbred females||Outbred females|
|Mother's ed. level (year)||6.04||2.79||9.00||3.89|
|Father's ed. level (year)||7.20||3.32||10.48||4.51|
|Head circumference (cm)a||34.84||1.00||35.16||0.83|
|Chest circumference (cm)a||33.25||0.85||33.56||1.03|
|Ear length FA (mm)a||1.24||1.67||0.99||1.09|
|Ear width FA (mm)a||1.00||0.94||0.88||1.07|
|Second digit FA (mm)a||0.77||0.65||0.44||0.76|
|Fourth digit FA (mm)a||0.38||0.58||0.41||0.45|
|Composite FA (mm)a||0.85||0.63||0.68||0.54|
We used a multivariate analysis of covariance (MANCOVA) to test the null hypothesis that the mean vectors of inbred and outbred newborns were sampled from the same population. Total years of parental education was the covariate. After removing the highly significant effect of education, first-cousin marriages had no detectable influence on the mean vectors of male and female newborns (Table 3).
Table 3. Multivariate analyses of covariance for male and female newborns
|Sex||Source of variation||Hypothesis DF||Error DF||Wilk's λ||F||P|
|Total years of education||5||92||0.755||5.959||<0.001|
|Total years of education||5||93||0.853||3.193||0.011|
As anticipated from the results of the MANCOVAs, first-cousin marriages had no detectable influence on any of the five dependent variables, in either males or females (Table 4). In contrast, all of the dependent variables in male newborns varied with total years of education, whereas only one dependent variable in females (weight) did so. More educated parents had larger male newborns (greater weight, taller stature, and larger head and chest circumferences) having less fluctuating asymmetry (Fig. 1). More educated parents had female newborns that weighed more.
Figure 1. Linear regression of CFA on total years of education, by sex and marriage type. Open circles and dashed lines are offspring of first cousins. Closed circles and solid lines are offspring of unrelated couples.
Download figure to PowerPoint
Table 4. Univariate analyses of covariance for male and female newborns
|Sex||Dependent variable||Marriage type||Total years of education||Error|
- Top of page
- MATERIAL AND METHODS
- LITERATURE CITED
When total years of education were taken into account as a covariate, inbred (f = 1/16) Turkish newborns in Sivas did not have greater fluctuating asymmetry than unrelated (f < 1/1,024) newborns. Moreover, they were of normal size at birth. Özener (2010), in contrast, found that inbred teens from urban Ankara had greater composite fluctuating asymmetry and were, on average, 3.52 kg lighter and 3.17 cm shorter than comparable outbred teens. While several previous studies have failed to detect greater fluctuating asymmetry of inbred populations (for reviews, see Leamy and Klingenberg, 2005; Graham et al., 2010), smaller body size is commonly observed in studies of inbreeding depression (Bittles and Black, 2010b) and the lack of it here requires an explanation.
The normal birth weight among the inbred infants may arise for several reasons. First, low weight may be an effect of inbreeding seen later in life. A more likely explanation, however, is that our study only examined full-term infants, with gestational ages between 37 and 41 weeks. We did not examine preterm infants, which almost certainly would have had lower birth weight (Kramer et al., 2001), and perhaps greater fluctuating asymmetry as well. In Jordan, preterm delivery accounts for most of the differences in birth weight between inbred (19.9% preterm) and outbred (12.3% preterm) infants (Obeidat et al., 2010). Similar results were obtained in Lebanon (Mumtaz et al., 2007; Mumtaz et al., 2010).
Because ours is an observational study, we cannot assign cause and effect with certainty. There may have been other, unobserved, variables influencing fluctuating asymmetry. For example, the parents of the inbred infants in our study had about 3 years less education than those of the outbred infants. Education may be confounded with other variables that influence size and asymmetry. If two populations in an observational study differ on one covariate (education), they must differ on others (Huitema, 1980). Smoking rates, for example, are greater among less educated individuals (Giskes et al., 2005). In Sivas, 4.9% of high school girls and 36.5% of high school boys smoke, and the percentage is about 10% higher for students whose parents have less than a high school education for fathers and less than a primary school education for mothers (Golbasi et al., 2011). Although education itself is probably causally unrelated to fluctuating asymmetry, it is almost certainly confounded with smoking and diet. Kieser et al. (1997), for example, have shown that dental asymmetry in humans is related to both smoking and obesity. Finally, when Obeidat et al. (2010) took smoking and other variables (i.e., age, body mass index, income, residency, smoking, parity, family history of low birth weight, preterm delivery, congenital anomalies, and medical problems during pregnancy) into account in a very large study of inbreeding in a Jordanian population (n = 3,269), some of the negative effects, such as low birth weight and increased number of still births, were no longer significant. Preterm deliveries and congenital malformations, however, were still higher in the inbred Jordanian group, even when smoking was taken into account.
Even if smoking and diet play no role in the relationship between consanguineous mating, education, size at birth, and fluctuating asymmetry, it is also possible that our results are an artifact of another kind of nonrandom mate selection. Lynch and Walsh (1998) have suggested that, in humans, low-quality (or low social status) parents may be more likely to marry other low-status individuals who may, or may not, be close relatives, simply because their range of prospective mates is limited. Denic and Agarwal (2012) describe just this scenario in a family that had lost social status for political reasons. If this is the case, then the observed inbreeding depression may be exaggerated by the expression of alleles (not necessarily recessive ones) associated with the low fitness. If low-status individuals are less likely to extend their education, and if low status has a genetic component, then the relationship between education, size at birth, and fluctuating asymmetry could also have a genetic component. This argument, however, makes more sense for societies in which first-cousin marriages are rarer than they are in central Turkey. Throughout Turkey, nearly 9% of the richest households feature first-cousin marriages (Koc, 2008). While this rate is less than half that of the poorest households, which have a rate of 22.3% in Turkey, it is still high by European standards. This then begs the question, why is inbreeding so prevalent in some societies and not in others?
Under some circumstances, consanguineous marriages may produce more children than nonconsanguineous pairings (Helgason et al., 2008; Bittles and Black, 2010a). And entire populations can benefit from consanguineous mating in places where malaria is endemic, by exposing alleles such as those associated with α+-thalassemia to positive selection (Denic et al., 2008, 2013; but see Bittles, 2011). This is not a new idea, however. Sewell Wright (1932, 1982) long argued that mutation, inbreeding, and selection interact to increase a population's rate of adaptation. This may be the situation in Turkey. Until recently, malaria was endemic in southeastern Turkey (Özbilgin et al., 2011; Piyal et al., 2013). This is an area where consanguineous marriages are common.
In conclusion, first cousin marriages in Turkey do not appear to increase the level of fluctuating asymmetry in newborns (this article), though they do so in urban teens (Özener, 2010). Moreover, first cousin marriages in Sivas do not cause size reduction of newborns, but this may be an artifact of selecting only full-term infants in our study. Future studies should also look into correlates of parental education, such as smoking behavior, diet, and exposure to environmental toxins.