Everything is (half) heritable (and half non-shared environment): "Meta-analysis of the heritability of human traits based on fifty years of twin studies", Polderman et al 2015 https://www.dropbox.com/s/ekxxlubswlry14k/2015-polderman.pdf / http://sci-hub.org/downloads/2575/10.1038@ng.3285.pdf(via http://infoproc.blogspot.com/2015/05/fifty-years-of-twin-studies.html )

Pretty astonishing work in its ambition - summarizing heritability from all twin studies. And the GCTAs say twin studies are right, so it can be interpreted pretty literally as genetics. Excerpts:

"Despite a century of research on complex traits in humans, the relative importance and specific nature of the influences of genes and environment on human traits remain controversial. We report a meta-analysis of twin correlations and reported variance components for 17,804 traits from 2,748 publications including 14,558,903 partly dependent twin pairs, virtually all published twin studies of complex traits. Estimates of heritability cluster strongly within functional domains, and across all traits the reported heritability is 49%. For a majority (69%) of traits, the observed twin correlations are consistent with a simple and parsimonious model where twin resemblance is solely due to additive genetic variation. The data are inconsistent with substantial influences from shared environment or non-additive genetic variation. This study provides the most comprehensive analysis of the causes of individual differences in human traits thus far and will guide future gene-mapping efforts. All the results can be visualized using the MaTCH webtool.

Specifically, the partitioning of observed variability into underlying genetic and environmental sources and the relative importance of additive and non-additive genetic variation are continually debated 1–5 . Recent results from large-scale genome-wide association studies (GWAS) show that many genetic variants contribute to the variation in complex traits and that effect sizes are typically small 6,7 . However, the sum of the variance explained by the detected variants is much smaller than the reported heritability of the trait 4,6–10 . This ‘missing heritability’ has led some investigators to conclude that non-additive variation must be important 4,11 . Although the presence of gene-gene interaction has been demonstrated empirically 5,12–17 , little is known about its relative contribution to observed variation 18 . In this study, our aim is twofold. First, we analyze empirical estimates of the relative contributions of genes and environment for virtually all human traits investigated in the past 50 years. Second, we assess empirical evidence for the presence and relative importance of non-additive genetic influences on all human traits studied. We rely on classical twin studies, as the twin design has been used widely to disentangle the relative contributions of genes and environment, across a variety of human traits.

Half of these were published after 2004, with sample sizes per study in 2012 of around 1,000 twin pairs (Supplementary Table 2). Each study could report on multiple traits measured in one or several samples. These 2,748 studies reported on 17,804 traits. Twin subjects came from 39 different countries, with a large proportion of studies (34%) based on US twin samples. The continents of South America (0.5%), Africa (0.2%) and Asia (5%) were heavily underrepresented (Fig. 1a,b and Supplementary Table 3).

The majority of studies (59%) were based on the adult population (aged 18–64 years), although the sample sizes available for studies of the elderly population (aged 65 years or older) were the largest (Supplementary Table 4). Authorship network analyses showed that 61 communities of authors wrote the 2,748 published studies. The 11 largest authorship communities contained >65 authors and could be mapped back to the main international twin registries, such as the Vietnam Era Twin Registry, the Finnish Twin Cohort and the Swedish Twin Registry (Supplementary Fig. 1).
The investigated traits fell into 28 general trait domains. The distribution of the traits evaluated in twin studies was highly skewed, with 51% of studies focusing on traits classified under the psychiatric, metabolic and cognitive domains, whereas traits classified under the developmental, connective tissue and infection domains together accounted for less than 1% of all investigated traits (Fig. 1c and Supplementary Tables 5–7). The ten most investigated traits were temperament and personality functions, weight maintenance functions, general metabolic functions, depressive episode, higher-level cognitive functions, conduct disorders, mental and behavioral disorders due to use of alcohol, anxiety disorders, height and mental and behavioral disorders due to use of tobacco. Collectively, these traits accounted for 59% of all investigated trait

We did not find evidence of systematic publication bias as a function of sample size (for example, where studies based on relatively small samples were only published when larger effects were reported) (Fig. 1d, Supplementary Figs. 2–6 and Supplementary Tables 8–11). We calculated the weighted averages of correlations for monozygotic (r MZ ) and dizygotic (r DZ ) twins and of the reported estimates of the relative contributions of genetic and environmental influences to the investigated traits using a random-effects meta-analytic model to allow for heterogeneity across different studies (Supplementary Tables 12–15). The meta-analyses of all traits yielded an average r MZ of 0.636 (s.e.m. = 0.002) and an average r DZ of 0.339 (s.e.m. = 0.003). The reported heritability (h 2 ) across all traits was 0.488 (s.e.m. = 0.004), and the reported estimate of shared environmental effects (c 2 ) was 0.174 (s.e.m. = 0.004) (Fig. 2a,b, Table 1 and Supplementary Fig. 7).

All weighted averages of h^2 across >500 distinct traits had a mean greater than zero (Supplementary Tables 17–24). The lowest reported heritability for a specific trait was for gene expression, with an estimated h^2 = 0.055 (s.e.m. = 0.026) and an estimated c 2 of 0.736 (s.e.m. = 0.033) (but note that these trait averages are based on reported estimates of variance components derived from only 20 data points reporting on the expression levels of 20 genes; Supplementary Table 21).

For the vast majority of traits (84%), we found that monozygotic twin correlations were larger than dizygotic twin correlations. Using the weighted estimates of r MZ and r DZ across all traits, we showed that, on average, 2r DZ − r MZ = 0.042 (s.e.m. = 0.007) (Table 1), which is very close to a twofold difference in the correlation of monozygotic twins relative to dizygotic twins (Supplementary Figs. 11 and 12). The proportion of single studies in which the pattern of twin correlations was consistent with the null hypothesis that 2r DZ = r MZ was 69%. This observed pattern of twin correlations is consistent with a simple and parsimonious underlying model of the absence of environmental effects shared by twin pairs and the presence of genetic effects that are entirely due to additive genetic variation (Table 2). This remarkable fitting of the data with a simple mode of family resemblance is inconsistent with the hypothesis that a substantial part of variation in human traits is due to shared environmental variation or to substantial non-additive genetic variation.

In only 3 of 28 general trait domains were most studies inconsistent with this model. These domains were activities (35%), reproduction (44%) and dermatological (45%) (Table 2 and Supplementary Table 27). Of the 59 specific traits (ICD-10 or ICF subchapter classifications) for which we had sufficient information to calculate the proportion of studies consistent with 2r DZ = r MZ , 21 traits showed a proportion less than 0.50, whereas for the remaining 38 traits the majority of individual studies were consistent with 2r DZ = r MZ (Supplementary Table 29). Of the top 20 most investigated specific traits, we found that for 12 traits the majority of individual studies were consistent with a model where variance was solely due to additive genetic variance and non-shared environmental variance, whereas the pattern of monozygotic and dizygotic twin correlations was inconsistent with this model for 8 traits, suggesting that, apart from additive genetic influences and non-shared environmental influences, either or both non-additive genetic influences and shared environmental influences are needed to explain the observed pattern of twin correlations (Table 2). These eight traits were conduct disorders, height, higher-level cognitive functions, hyper-kinetic disorders, mental and behavioral disorders due to the use of alcohol, mental and behavioral disorders due to the use of tobacco, other anxiety disorders and weight maintenance functions. For all eight traits, meta-analyses on reported variance components resulted in a weighted estimate of reported shared environmental influences that was statistically different from zero (Supplementary Table 21). Comparison of weighted twin correlations for these specific traits resulted in positive estimates of 2r DZ − r MZ , except for hyperkinetic disorders, where 2r DZ − r MZ was −0.130 (s.e.m. = 0.034) on the basis of 144 individual reports and 207,589 twin pairs, which suggests the influence of non-additive genetic variation for this trait or any other source of variation that leads to a disproportionate similarity among monozygotic twin pairs."