Bringing genetics back to psychiatric endophenotypes.

t T he concept of endophenotypes, introduced almost 4 decades ago, has become increasingly important in the study of complex neuropsychiatric diseases. Endophenotypes are easurable, but not overtly observable, constructs in the pathway rom genetic variation to psychiatric disorder. As originally concepualized, endophenotypes must: 1) be associated with illness in the opulation; 2) be heritable; 3) be state independent, although they ay need to be elicited by a challenge; 4) cosegregate with illness ithin families; and 5) be found in unaffected relatives of probands t a higher rate than in the general population (1). Initially, scientists ad hoped that investigating endophenotypes would hasten the dentification of the genetic origins of various disorders, but this has ot proven to be the case. This difficulty in identifying the genetic oundations of psychiatric disorders is due, in part, to the restrictive ature of the definitions included in the International Classification f Diseases, 10th revision, and the Diagnostic and Statistical Manual f Mental Disorders-IV, which are not based on etiology or pathohysiology (2). But it is also due to the fact that investigators have eparted from the original criteria for endophenotypes noted bove and have considered as a possible endophenotype virtually ny neural, cognitive, or affective abnormality associated with a ental illness, regardless of its genetic links. This overgeneralizaion has impeded the study of endophenotypes in promoting the iscovery of genes underlying psychopathology. In this issue of Biological Psychiatry, Glahn et al. (3) present work hat both returns to the original conceptualization of endophenoypes (with, arguably, one notable exception described below) and rovides a formal, mathematical procedure for identifying endohenotypes that may promote gene discovery in psychiatry. In the ontext of major depressive disorder (MDD), these researchers dene and test the usefulness of the endophenotype ranking value ERV), an index for measuring the strength of association with geetics of endophenotypes implicated in psychiatric illness. The ERV s a straightforward index that increases with increasing values of hree constructs: 1) the heritability of the psychiatric illness; 2) the eritability of the endophenotype; and 3) the genetic correlation etween the illness and the endophenotype. In this formulation, a utative endophenotype that is associated with a heritable disorer and that is itself heritable will receive a high ERV, provided that t has high genetic correlation with the disorder; thus, the ERV eturns genetic linkage to the conception of the endophenotypes nd also maintains the requisite links between the endophenotype nd the illness. Glahn et al. (3) calculated ERVs for a large set (more than 11,000) f behavioral, neural, and transcriptomic factors that were assessed n a sample of 1122 Mexican-American persons. They then perormed quality tests on the ERV index by conducting univariate and ivariate linkage analyses of the endophenotypes from each of hese three categories that have the highest ERVs: the Beck Depresion Inventory (BDI) scores, the bilateral diencephalon volume, and he RNF123 transcriptional endophenotype, as well as the inci-