Afterword

Gottesman and Gould's 2003 paper on endopheno-types¹ as a research strategy for psychiatry has almost 800 citations and the present volume is an impressive summary of how far this field has come. Recently, whole genome wide scans for schizophrenia and bipolar disorder with excellent methodology and large, well diagnosed samples have yielded few replicable associations, and those found were of small effect size.² These results make the endophenotype strategy even more relevant than it has been up to now. It seems likely that we will first fi nd genes for components of the processes of mental illness and behavior and only later find how these components interact to produce normal and abnormal behavior.

It is a source of pride for our field that the most up to date methodology in biology, molecular biology, proteomics and imaging is being actively applied to the challenge of understanding psychiatric disorder. The editor and authors of this volume are optimistic and indeed a degree of optimism and sense of mission is essential for science. On the other hand, all the authors are appropriately guarded as to the usefulness in clinical practice of any fi ndings up to now and emphasize the limitations of using particular peripheral markers distant from the brain or derived variables such as electrophysiology for understanding mental illness. The endophenotype concept is now beyond its childhood and well into its adolescence but it has not yet led to the unraveling of the complex problems we face in understanding psychiatric disorder. It is relevant here that the several chapters in this volume on Alzheimer's, a disease with a clear postmortem pathology and neurochemistry and marked imaging fi ndings in the later stages, has also not yielded to endopheno-type approaches and we have no early biomarkers that could allow us to counsel persons at risk for Alzheimer 's or apply early preventive treatments.