Abstract

In lieu of an abstract, here is a brief excerpt of the content:The Post-Genomic RevolutionA Paradigm Shift for Biopsychosocial SystemsClaude Robert Cloninger, MD, PhD (bio)The pstchologist Danielle Dick and psychiatrist Kenneth Kendler (DK) began an ongoing study in 2011 called Spit for Science (S4S) in which they obtained saliva as a peripheral source of DNA along with assessment of detailed self-report information on alcohol and other substance use, selected personality traits, and psychosocial history about the students entering a large university (Dick et al., 2014). They hoped to identify the genes underlying the heritability of alcohol use and related behaviors to inform students about their individual risk, and thereby to promote increased self-awareness and health. The students were highly cooperative volunteers, and many peer-reviewed research papers have been published from these data.Eric Turkheimer, a psychologist who studies how genes and environments shape the development of human behavior (Harden et al., 2007; Loehlin et al., 2009; Turkheimer et al., 2003), reviewed all the resulting publications coming from S4S along with coauthor Sarah Greer (Turkheimer & Greer, 2024). Turkheimer and Greer (TG) were surprised to find the study has failed to identify any genes that have substantial (“nontrivial”) effects on reported levels of consumption or problems related to their use of alcohol or other substances, except for a variant of the gene for aldehyde dehydrogenase (ALDH2), which was already known to cause many people of Asian descent to flush when they drink alcohol. TG state that, “Again and again, the results of S4S investigations demonstrated tiny, and frequently null, contributions of genetics to differences in drinking and related behaviors among college students. GWAS [genome-wide association studies] revealed practically no genome-wide significant SNPs [single nucleotide polymorphisms], and the SNPs identified at less stringent levels of significance were never mentioned more than once” (Turkheimer & Greer, 2024).TG concluded that the design, execution, psychometric measures, and analysis of the data were excellent. However, they also discuss facts that raise doubts about the adequacy of the design and methods. Namely, it is an inconvenient fact, [End Page 429] but widely known, that approaches to GWAS that are frequently used in psychiatric genetics require huge samples (often hundreds of thousands) to detect statistically significant genes because the observed effects have consistently been weak whenever genes are assumed to act independently, as did DK. Consequently, polygenic risk scores are weak and inconsistent predictors of behavioral phenotypes (Zwir et al., 2020c). Nevertheless, TG fault DK and their coauthors for an overemphasis on statistical significance in large samples rather than prominently reporting that the observed effect sizes of genes were only weakly and inconsistently associated with any measured features of substance use.TG report that they were surprised by the weak and inconsistent findings from genetic variants from this GWAS because they knew that twin and family studies (i.e., what they call “genetics at altitude”) have previously found strong and consistent evidence of the heritability of phenotypes like those measured by DK. For example, in my own research with adoptees separated from their biological parents at an early age and reared by adoptive parents, the risk of alcohol abuse in adoptees was strongly predicted by alcohol abuse and related behaviors in their biological, but not adoptive, parents (Bohman et al., 1981; Cloninger et al., 1981; Sigvardsson et al., 1996) and by their childhood personality, particularly novelty seeking, when the children were followed into young adulthood (Cloninger, 1987; Cloninger et al., 1988). However, it is well-known that the methods used by DK for GWAS consistently fail to detect most of the heritable variance for behavioral phenotypes estimated in twin studies, which include much variability from complex patterns of gene–gene interaction (Table 1) (Cloninger & Zwir, 2022). This consistent observation in GWAS studies that assume genes act independently is called the missing or hidden heritability problem.Nevertheless, DK expressed optimism in 2011 that genome-wide measurement of genetic variants would identify genetic markers to predict risk of alcohol-related phenotypes in college students. They do so because they assume that all the information about heritable processes can be extracted from genome-wise scans of SNPS. The heritable phenotypes of interest in S4S included patterns of use of alcohol and other recreational...