Digit ratio (2D:4D) and body mass index in the BBC Internet Study: prenatal sex steroids and a Trivers-Willard effect on body composition.

Digit ratio - a putative measure of prenatal sex steroids - may be related to body mass index (BMI). However, reports of correlations between 2D:4D and BMI have yielded mixed results with some studies showing no relationship while others have reported positive associations in men or women only. This study considers associations between self-reported 2D:4D and BMI in a large online survey (i.e. the BBC Internet Study). At the individual level, there was a weak positive association between 2D:4D and BMI in both sexes with greater effect sizes in women. Body mass index was positively related to age and negatively related to parental income; however, the relationship between 2D:4D and BMI was independent of both variables. At the national level, mean 2D:4D per country showed positive associations with mean national BMI but those correlations were restricted to females. It is concluded that BMI is positively related to low prenatal testosterone and high prenatal oestrogen. Parental income inequality may influence both prenatal sex steroids (through a 'Trivers-Willard' effect) and BMI such that increases in inequality result in reductions in prenatal testosterone and increases in BMI at the individual and national level.

The distribution of B chromosomes across species.

In this review we look at the broad picture of how B chromosomes are distributed across a wide range of species. We review recent studies of the factors associated with the presence of Bs across species, and provide new analyses with updated data and additional variables. The major obstacle facing comparative studies of B chromosome distribution is variation among species in the intensity of cytogenetic study. Because Bs are, by definition, not present in all individuals of a species, they may often be overlooked in species that are rarely studied. We give examples of corrections for differences in study effort, and show that after a variety of such corrections, strong correlations remain. Several major biological factors are associated with the presence of B chromosomes. Among flowering plants, Bs are more likely to occur in outcrossing than in inbred species, and their presence is also positively correlated with genome size and negatively with chromosome number. They are no more frequent in polyploids than in diploids, nor in species with multiple ploidies. Among mammals, Bs are more likely to occur in species with karyotypes consisting of mostly acrocentric chromosomes. We find no evidence for an association with chromosome number or genome size in mammals, although the sample for genome size is small. The associations with breeding system and acrocentric chromosomes were both predicted in advance, but those with genome size and chromosome number were discovered empirically and we can offer only tentative explanations for the very strong associations we have uncovered. Our understanding of why B chromosomes are present in some species and absent in others is still in its infancy, and we suggest several potential avenues for future research.