Headlines often create the message. I asked Brian Koehler to comment on this editorial in the latest Am J Psych and specifically the sentence in para 2 below, "Indeed, the majority of the susceptibility for certain psychiatric disorders, such as bipolar disorder and schizophrenia, is due to inheritance (5)." See reference Generously, despite getting ready for the ISPS USA conference at the end of this week, Brian promptly replied as follows "This has not been unequivocably demonstrated. See below:

Sullivan and colleagues (2006-"Textbook of Schizophrenia" edited by Jeffrey A.Lieberman, T. Scott Stroup and Diana O. Perkins for American Psychiatric Publishing, Inc.) pointed out that the "pathogenesis of schizophrenia is unknown, and no compelling biological markers of sufficient sensitivity and specificity exist" (p. 40). "

This is complex territory about models of science and how models are presented in simplified form, and how statistical models may inform or mislead the clinical care of individual patients.

Brian's futher comments leading on from the quote above, are below, followed by the editorial in the latest American Journal of Psychiatry, for those interested in unpacking this important subject.

Chris Burford

Brian continued from the quote above:-

"However, they noted that current research on the molecular genetics of schizophrenia nominates several genes of potential importance in the etiology of schizophrenia: neuregulin 1 (NRG1), dystrobrevin binding protein 1 (DTNBP1), G72 and G30, regulator of G-protein signalling 4 (RGS4), catechol-o-methyltransferase (COMT), proline dehydrogenase (PRODH), disrupted in schizophrenia 1 and 2 (DISC1 and DISC2), serotonin 2a receptor (HTR2A), and dopamine 3 receptor (DRD3). The evidence for etiological risk is strong for genes NRG1 (although a recent study published in Psychological Medicine, 2005, 35, 1599-1610, "Neurgulin 1 (NRG1) and schizophrenia: analysis of a US family sample and the evidence in the balance," failed to find an association between NRG1 and schizophrenia) and DTNBP1, intermediate for genes RGS4, G72/G30/DAO, HTR24, and DRD3 and weak for genes COMT and PRODH. It is important to bear in mind that one should be cautious in the etiologic interpretation of the results of linkage and association studies-the classical molecular genetic study designs in humans-because they only yield statistical support for candidate genes for schizophrenia. These studies alone cannot prove the etiological significance of a particular gene and it is likely that some of these candidate genes will prove to be false positives. However, it would be a monumental advance if even one of these genes prove to be of etiological significance.

The contribution of individual genes to the etiology of schizophrenia seems quite small. Odds ratios (ORs) for recently reported susceptibility genes are less than 2.0. Gilmore and Murray (2006) point out that the OR for the neuregulin (NRG) gene is in the 1.25 to 2.20 range. For the catechol-O-methytransferase (COMT) gene, the OR is at best 1.59. For the dysbindin gene the OR is 1.14 to 1.87. The ORs for several environmental risk factors are similar to those of individual candidate genes in schizophrenia. Environmental risk factors such as urban birth and certain obstetrical complications, e.g., premature rupture of membranes, placental abruption, emergency cesarean delivery and hypoxia-ischemic-related complications, have actually higher ORs than any individual candidate gene.

Gilmore and Murray (2006) noted that twin studies estimate the heritability of schizophrenia to be approximately 80%, with the remaining 20% variance to be attributed to environmental factors. Concordance rates in MZ twins are only in the 40%-45% range, and some researchers have good reason to feel this is an inflated figure. The twin methodology, according Structural Neuroimaging Research in Schizophreni to Gilmore and Murray, has two major limitations which may lead to overestimating the genetic contribution and underestimating the environmental contribution to schizophrenia. First, heritability estimates assume additive effects of genes and environments and do not allow for gene-environment interactions. Secondly, twin methodology is based on an equal environment assumption, i.e., that MZ and DZ twins share similar prenatal environments. This assumption is false since most MZ twins share a placenta, whereas DZ twins have separate placentas. In addition, MZ twins have lower birth weights, higher rates of premature birth, and more adverse perinatal outcomes than do DZ twins. Gilmore and Murray suggest: "These same pre-and perinatal environmental risk factors also may increase risk for schizophrenia and could therefore lead to higher concordance rates in MZ twins that would be falsely attributed to genetic factors" (p. 55). Some researchers have pointed out that the results of twin studies would be particularly misleading in disorders in which the prenatal environment plays a role in their etiology. Research has demonstrated that this may be the case in schizophrenia. Prenatal stress mediated by corticosteroids, results in a variety of abnormalities in the developing brain. Of particular relevance to schizophrenia, prenatal stress results in decreased synaptophysin immunoreactivity in the cortex and hippocampus of adult animals (Hayashi et al 1998; Koo et al 2003), as well as alterations in pre- and postsynaptic gene expression in frontal cortex (Kinnunen et al 2003). Prenatal stress inhibits neurogenesis in the hippocampus (Coe et al 2003; Lemaire et al 2000). Hippocampal GABA systems are also altered by prenatal stress (Stone et al 2001) and prenatal stress results in long-lasting alterations in the neuroendocrine response to stress in animals, which also could contribute to ongoing central nervous system dysfunction in mammals (Edwards and Burnham 2001; Matthews 2000).

I shall conclude this section on genetics with observations by Richard Lewontin (2000), Professor at Harvard University, taken from his volume "The Triple Helix: Gene, Organism, and Environment" (published by Harvard University Press):

"The reigning mode of explanation at present is genetic. Reinforced by the observation that some human disorders result from mutation of clearly defined genes, nearly all human variation is now ascribed to genetic differences. From the undoubted fact that gene mutations like the Tay-Sachs mutation or chromosomal abnormalities like the extra chromosome causing Down syndrome are the sources of pathological variation, human geneticists have assumed that heart disease, diabetes, breast cancer, and bipolar syndrome must also be genetic variants. The search for genetic variation is a major preoccupation of medical research...The trouble with the general scheme of explanation contained in the metaphor of development is that it is bad biology. If we had the complete DNA sequence of an organism and unlimited computational power, we could not compute the organism, because the organism does not compute itself from its genes....There exists, and has existed for a long time, a large body of evidence that demonstrates that the ontogeny of an organism is the consequence of a unique interaction between the genes it carries, the temporal sequence of external environments through which it passes during its life, and random events of molecular interactions within individual cells. It is these interactions that must be incorporated into any proper account of how an organism is formed "(pp. 17-18).

Lewontin concluded:

"The organism is determined neither by its genes nor by its environment nor even by the interaction between them, but bears a significant mark of random processes. The organism does not compute itself from the information in its genes nor even from the information in the genes and the sequence of environments. The metaphor of computation is just a trendy form of Descartes's metaphor of the machine. Like any metaphor, it catches some aspect of the truth but leads us astray if we take it too seriously" (p. 38).

In regard to genes and mental disorder in generals, I would summarize that the current thinking by geneticists is as follows:

Schizophrenia, or mental illness in general, may be thought to arise from polygenic factors, i.e., many genes, of low to moderate effect, which are in a dose-response non-linear relationship with multiple hgh risk environmental factors. Neurogenetic reductionism in psychiatry, is what it has always been, a costly and erroneous short-cut to the comprehensive understanding and treatment of complex mental disorders which arise from complex non-liinear interactions between genes (including epigenetic processes), brain, mind (which includes multiple factors at the level of the individual and family), as well as the sociocultural surround (a complex framework which includes vertical and horizontal dimensions, especially involving the transgenerational transmission of trauma, or what Finnish psychiatrist-psychoanalyst Martti Siiral calls the transfer of burden which is waiting for the chance of dialogical encounter, perhaps in what psychoannalysts have called the transference).

In conclusion, for those interested in current thinking on the role of genes in mental disorders, and in the human condition generally, I would highly recommend Sir Michael Rutter's "Genes and Behavior: Nature-Nurture Interplay Explained" published in 2006 by Blackwell Publishing.

Brian Koehler PhD
Postdoctoral Program
New York University
80 East 11 Street #339
New York NY 10003
brian_koehler@psychoanalysis.net
212.533.5687