Everything is heritable; "A Genome-Wide Analysis of Liberal and Conservative Political Attitudes", Hatemi et al 2011  (apparently I missed this when it came out); excerpts:

"The assumption that the transmission of social behaviors and political preferences is purely cultural has been challenged repeatedly over the last 40 years by the combined evidence of large studies of adult twins and their relatives, adoption studies, and twins reared apart. Variance components and path modeling analyses using data from extended families quantified the overall genetic influence on political attitudes, but few studies have attempted to localize the parts of the genome which accounted for the heritability estimates found for political preferences. Here, we present the first genome-wide analysis of Conservative-Liberal attitudes from a sample of 13,000 respondents whose DNA was collected in conjunction with a 50-item sociopolitical attitude questionnaire. Several significant linkage peaks were identified and potential candidate genes discussed.

There is marked family resemblance for political attitudes (Alwin, Cohen, and Newcomb 1991; Jennings and Niemi 1982). The dominant paradigm in the social and political sciences has asserted that this resemblance is due almost entirely to social learning, social background, or environmental influences (Campbell et al. 1960; Converse 1964; Jennings and Niemi 1968). Indeed, social and political behaviors have been used to illustrate elegant mathematical theories of cultural inheritance (Cavalli-Sforza and Feldman 1981). However, over the last 40 years, a series of very large studies of twins, families, and adoptions have given strong support to the alternative view that a significant component of family resemblance on political attitudes is genetically influenced (Eaves, Eysenck, and Martin 1989; Eaves et al. 1999; Martin et al. 1986). More recently, political scientists have also adopted a ''partially genetic'' approach (Alford, Funk, and Hibbing 2005; Fowler, Baker, and Dawes 2008; Hatemi et al. 2009) and using longitudinal and extended pedigree studies further elaborated on the genetic influence for political attitudes (Eaves and Hatemi 2008; Hatemi et al. 2010).
...identifying genes that influence the development of neurological functions and release of hormones, which in turn influence cognition and interpretation of environmental stimuli, could indicate whether the genetic basis of political preferences share common biological influences with cognition (Gazanniga 2004), reactions to threat (Oxley et al. 2008), morality (Lakoff 2002), disgust (Rozin, Haidt and Fincher 2009), emotion reading (Dolan 2002), risk taking (McClure et al. 2004), reproductive differences (Skuse 2005), fear (Hatemi and McDermott 2009), aggression (McDermott et al. 2009), pursuit of power (Madsen 1986), or rational action (Sanfey et al. 2003), to name just a few possibilities.

There are two basic approaches to identifying genes for complex traits. The first exploits a priori information that suggests a certain gene might be expected to be associated with a specific trait. This ''candidate gene'' approach has recently been undertaken for political traits and several genetic markers have been implicated using allelic 2 association methods, including monoamine oxidase and serotonin for voting participation (Fowler and Dawes 2008), serotonin for certain attitude positions (Hatemi et al. 2008), and dopamine for ideology (Dawes and Fowler 2009). However, these candidate gene studies were undertaken because those genes were previously found to be associated with traits similar to political attitudes. This is important because in most studies of complex human traits, candidate gene studies not supported by genome wide methods have typically been unrewarding and not replicated. For example, in possibly the most celebrated of allelic gene-environment interaction studies, Caspi et al. (2003) reported that 5-HTTPLR (serotonin) moderates the effects of stressful life events on depression. However, a recent meta-analysis provided evidence that Caspi's results do not withstand empirical validation (see Risch et al.2009).
As a result, geneticists have tended to pay more attention to a second approach; to search systematically across the genome for genes that demonstrate significant association with the trait in question while taking into account all other identified genetic markers. There are over 20,000 identified genes in the human genome, many of which we know little about. Genome-wide methods allow researchers to search for individual genetic markers or chromosomal regions that influence a trait, regardless of their function, and offer a much more robust test than candidate gene studies which are biased by the choice of genes included. Thus, genome-wide analyses can implicate specific genes or regions of genes that we did not suspect were influencing a trait of interest and thus reveal novel pathways to the formation of political orientations. Although there is a growing volume of evidence confirming Fisher's (1918) conjecture that human physical and behavioral traits reflect the cumulative small effects of a very large number of individual genetic loci (e.g., Benyamin et al. 2008), so far, there has been no systematic empirical attempt to begin such analysis for individual differences in political preferences using sample sizes that come near those required to detect the relatively small effects of individual quantitative trait loci (QTL) 3 likely to account for the genetic components of complex traits. Here, on over 13,000 individuals, we present the results from the first genome-wide analysis on Conservative-Liberal orientations which identifies several significant genetic focal points.

All participants were relatives of Caucasian ancestry. Participants completed a Health and Lifestyle Questionnaire (HLQ) between 1988 and 1990 and gave blood samples for DNA extraction and genotyping. Conservatism-Liberalism was assessed by a 50-item scale of contemporary socially and politically relevant Australian attitudes designed to be similar to the Wilson-Patterson (1968) inventory. The inventory presented participants with a short stimulus word or phrase and asked them to respond positive, negative, or neutral to each (the specific attitudes and question format are presented in Appendix B). Confirmatory factor analysis of these items on this population has shown that a uni-dimensional model fits the data, but three to five correlated sub-factors may also be extracted (Martin et al. 1986; Verhulst, Hatemi, and Martin 2010). The present analysis focuses on the first general factor of Conservatism-Liberalism which is normally distributed (Appendix C) and widely used in the extant literature (Bouchard et al. 1990). Complete Conservatism-Liberalism scores were available on 20,725 individuals from 8,139 families, which included parents, siblings, spouses, and offspring. However, the number of participants for which we have genotypic data is less. Complete phenotypic and genotypic data were available for 13,201 people from 2,774 families (for family structure see Table 1).

In order to detect which chromosomal regions may be responsible for variation in our Conservatism Liberalism factor we conducted variance components genome-wide linkage analysis with age and sex as covariates in MERLIN 1.1.2 (Abecasis et al. 2002) using a 10cM grid with maximum likelihood estimated allele frequencies. This commonly used method tests for co-segregation of chromosomal regions and the phenotypic trait of interest (Conservatism-Liberalism). The significance of each individual marker is assessed by comparing the difference in the log 10 likelihood between a model that includes the marker and one in which the marker's effect is fixed to zero. The logarithm of odds (LOD) score is used as a guide to assess the weight of evidence in favor of linkage at each location throughout the genome and is the statistical test for determining significant linkage. LOD scores greater than 3.0 are significant (Lander and Kruglyak 1995) and the likelihood of observing a significant genetic polymorphism which is not linked to the trait is less than 1 in 1000. However, linkage analysis has limited power to detect genes of small effects. When many genes contribute to trait variation, as is almost certainly the case with political temperament, then even very large linkage studies may not have sufficient statistical power to detect the effects of most genetic loci (Carey 2003). Thus, it is conventional in genetic studies to report LOD scores above 2.5 as ''suggestive'' and pursue replication in future studies (e.g., Bakker et al. 2003).

Linkage signal regions are large and can include many genes. Linkage does not identify a specific genetic marker, but regions on a chromosome.

Figure 1 plots the LOD scores by location for the genome-wide variance component linkage analysis of the Conservatism-Liberalism attitude factor, adjusted for age and sex. Three of the peaks reach genome-wide significance (LOD .3) as defined by Lander and Kruglyak (1995) and one peak reaches ''suggestive'' linkage (LOD .2.5)...Significant LOD score peaks were found on: chromosome 4 between 137,000K and 178,000K base pairs (bp) with the peak at 142,000K bp (LOD score of 3.44); chromosome 9 between 135,000K and 165,000K bp with the peak at 150,000K bp (LOD score of 3.09); and chromosome 2 between 130,000K and 155,000K bp with the peak at 142,000K bp (LOD score of 3.01). One suggestive LOD peak was found on chromosome 6 between 78,000K and 98,000K bp with the peak at 93,000K bp (LOD score of 2.55). These four most significant QTLs accounted for an estimated 12.9%, 12.9%, 12.7% and 9.4% of the total phenotypic variation on the Conservatism-Liberalism attitude factor.

[These are remarkably large effect sizes, but this is a consequence of their relatively small sample size: only large effect sizes will reach statistical-significance with their relatively small n=13k.]

As we identified four regions of interest, and one that meets the strictest criteria, our findings are consistent with what might be expected if the genetic component of variation in Conservatism-Liberalism resembles any other polygenic human trait, for which the genetic resemblance between relatives can only be resolved reliably into the effects of a large number of genes with small effects that typically cannot be identified by linkage. This makes our findings all the more intriguing. That is, the expectation that we should find no significant linkage peaks for such a complex human trait as political orientation is not supported. Rather, we find the opposite, which is somewhat remarkable, because for many biometric traits, more commonly viewed as being genetically influenced, such as height, LOD peaks are most often well below 3, as most genetic influences are the cumulative effect of a very large number of genes with very small effects, and linkage signals seldom have the power to identify such small effects. Here we found regions of interest which contain many genetic loci, but only a few loci in those regions have been previously identified as having some relationship with human social behavior. While the findings offer reasons for optimism, the relatively modest LOD scores are consistent with what we know about genetics at this point in time and what we should expect for such a complex human trait.

This being said, of the 60 known genetic markers within the 95% CI of the LOD peak (142,000K) on Chromosome 4, only one, which was also located a minimal distance from the peak, has previously been associated with human social behavior (see online Appendix E). NARG1, an N-methyl-D-aspartate (NMDA) receptor is located at 140,442K bp and has a significant LOD score of 3.38. NMDA is part of the ionotropic family of glutamate receptors. Glutamate is the principal excitatory neurotransmitter in the brain and is directly involved in a wide array of cognitive functions such as memory and learning (Jamain et al. 2002). The gene NARG1 is understood to encode an N-acetyltransferase protein thought to be important for vascular, hematopoietic, and neuronal growth and development (Sugiura, Patel, and Corriveau 2001). It is expressed at high levels in the testis and ocular endothelial cells, but also found in the largest connective pathway in the human brain (corpus callosum). NMDA receptors have been found to play an important role in a wide range of physiological, behavioral, and cognitive processes in mammals and contribute to synaptic transmission at sites throughout the brain and spinal cord. Both human and animal studies have identified NMDA being related to cognitive-behavioral performance, working memory, counting behavior, social learning, fear conditioning, spatial learning, motor performance, and social interaction, to include prosocial, antisocial, and aggressive behaviors (Duncan et al. 2004; Gewirtz and Davis 1997).

It makes sense to also include the ''suggestive'' serotonin receptors on chromosome 6 in this discussion with NMDA and political orientations, as both NMDA and serotonin have been significantly associated with a similar array of traits relevant to human social behaviors. Indeed, in various neurochemical studies, functional relationships between serotonin and NMDA have been reported (Madden and Morrison 2006; Shishido et al. 2000; Yuen et al. 2005) as they are both critically involved in the regulation of cognition and emotion (Canli and Lesch 2007; Hariri and Holmes 2006). Among the more intriguing relationships are the role both NMDA and serotonin have as regulators of fear, stress, and anxiety (Dai et al. 2008; Hariri et al. 2002; Young et al. 2007). Recently, there has been renewed interest in the influence of threat, fear and anxiety on ideological positions (Lupia and Menning 2009). For example, Oxley et al. (2008) found physiological differences in threat reaction between those with more conservative positions on outgroups (e.g., attitudes on immigration), while Hatemi and McDermott (2009) found that individual differences in fear dispositions were a significant predictor of political attitudes, and that the relationship between social fear and outgroup attitudes was largely of function of shared genetic influence (also see Jost et al. 2008). Furthermore, certain political dispositions and behaviors, such as political participation (Fowler and Dawes 2008) and power seeking (Madsen 1986) have been found to be significantly related to serotonin receptor length and whole blood levels of serotonin respectively. Based on the combination of our findings with those in the extant literature, the NMDA receptor NARG1, as well as KYNU and the serotonin loci HTR1E and HTR1B, are promising genes for allelic relationships with political attitudes, particularly those that are influenced by fear or anxiety. Future explorations would benefit by also including associated downstream neurobiological pathways, to include hormone regulation.

Although the current sample is the largest used so far in any attempt to identify specific genetic influences on social and political attitudes, and the only sample implicating individual QTLs contributing to individual differences in political attitudes, it still has low power to detect genes which have very small effects that typically account for the vast majority of genetic variation in complex behaviors (Fulker and Cherny 1996). Nevertheless, our largest chromosome-region effects explain ~13% of the total variance, implying that the gene accounting for this QTL is substantially correlated (sqrt(0.13) = 0.36) with Conservatism-Liberalism. However, we recognize the larger estimates of QTL effects in genome scans are typically biased upwards and our simulations show that estimates of 8% are not unusual even if there were no true
linkage.

Do political temperaments spring from the same biological sources as other personal temperaments and psychological traits, or is political thought distinct from other components of our behavior? Which biological systems are connected to which political belief structures? While, at all costs, we want to avoid claiming too much for early findings, our data give preliminary support to the hypothesis that whatever relationship exists between politics and genetics, it may be those genetic loci that influence flexibility in information processing and cognition. There is also some evidence, though weak, that the biological systems which influence political attitudes may be the ones related to those which regulate fear and anxiety (Hatemi and McDermott 2009; Oxley et al. 2008) or even possibly mate selection and disgust (Eaves et al. 2010; Navarrete and Fessler 2006). Whether or not it is a function of fear and loathing, betrothing and sexual desire, success of offspring, or other factors, eventually we may better understand the genetic variance behind political dispositions through locating genes by genome-wide analyses and working through the biological mechanisms that those genes are known to influence. We contend the pursuit of such knowledge is best approached using a variety of neurobiological, cultural, and environmental methods.
Taking care to avoid overstatement in advance of replication, the finding that glutamate and NMDA receptors are located on every significant or suggestive chromosomal region related to Conservative-Liberal attitudes provides reason to explore a previously uncharted pathway to how ideologies are formed. Future studies, directly exploring glutamate and NMDA's role in information processing, attitude formation and constraint, particularly during critical neurological development in childhood, which corresponds to the same critical period of social learning and cultural assimilation of values, may offer a better understanding of political ideology. Indeed, NMDA's function in learning and memory during development is of primary interest for future study."

#genetics #politics