In the past decade, several studies have highlighted the importance of differential genetic and environmental effects on a variety of traits across males and females. For example, a study performed in 806 subjects from a genetically isolated Hutterite population showed sex differences in X-linked and autosomal additive genetic effects on several anthropometric traits, such as height, fasting insulin, and triglycerides. Similarly, several robust sex-specific genetic effects have been identified outside the sex chromosomes in human³ and animal studies. In addition, sex-specific environmental effects have been observed in obesity, where boys and girls differ in susceptibility to their social environment. The presence of sex-specific etiological effects in human traits has far-reaching implications. A strong contribution of sex-specific effects in trait variation would for instance imply a difference in etiology between males and females for the same disorder, which would require sex-specific treatments. To assess the overall importance of sex-specific genetic effects across human traits, estimates of sex-specific heritability and male-female genetic correlations are most informative. While heritability quantifies the relative contribution of genetic effects compared to environmental effects in a trait, male-female co-heritability quantifies to what extent the same genetic variants play a role in males and females. Both measures are largely independent and complementary, and can be assessed using pairs of monozygotic (MZ) and dizygotic (DZ) twins, including same-sex and opposite DZ twins.

Twin studies have contributed enormously to our understanding of the relative contribution of genetic and environmental effects across human traits. A recently published meta-analysis of virtually all twin studies provided sex-specific estimates for the contribution of genes and environment across all traits investigated thus far. However, this study did not include an in-depth analysis of the extent to which male and female heritability estimates differed, nor did it report on the co-heritability in males and females. In the present study we therefore use single-sex and opposite-sex twin correlations to systematically assess the overall contribution of sex-specific genetic effects as well as the male-female genetic overlap across all investigated domains of human traits.

The use of twin correlations allows testing of three sex-specific hypotheses. The first hypothesis is that the difference between the monozygotic twin correlation and dizygotic twin correlation is similar across males and females . This can be interpreted as a test of whether the influence of (additive) genetic effects on the population variance is the same in males and females hypothesis of equal amount of heritability. The second hypothesis is that the difference between and is similar across sexes, thus . This can be interpreted as a test of whether the relative influence of the shared environment is the same in males and females (hypothesis of equal amount of shared environmental influences) The third hypothesis that can be tested with twin data is that the observed correlation in DZ same-sex twins here defined as the male and female DZ twin correlation midpoint is the same as the observed correlation in DZ twins of opposite sex. This is a test of whether the same genes and regulatory regions (and/or shared environmental factors) affect the trait of interest in males and females (hypothesis of full co-heritability).

Results

To assess the contribution of sex-specific genetic effects we used the MATCH database of twin correlations of Polderman of which we selected all traits for which sex-specific twin correlations and their standard errors were available. The three hypotheses as introduced above were tested per ‘trait’ as reported in the original study as well as per ‘trait category’ following the official ICF and ICD-10 sub-chapter classification .We report on 50 trait categories based on 2,608 reported traits to test our hypotheses