Perspectives in Psychiatric Genetics

In Perspectives in Neuropsychiatry, ed. D. Richter, pp. 173‑82. London: H. K. Lewis.



   Progress in science is very closely connected with the appearance and development of techniques of investigation, and with their abandonment when newer and better methods are found. So it has been in psychiatric genetics. Its beginnings are lost in the mists of time, when clinicians gained from a study of their patients the impression that mental dis­order was unusually frequent among the blood‑relatives of the mentally disordered. The prescientific age was dominated by the notion of polymorphism in psychiatry. The heredity of mental illness was believed to follow the theories of degeneration and anticipation, which inspired not only works of science but also of art, such as Samuel Butler's Way of All Flesh and Thomas Mann's epic Buddenbrooks. When the frequency of mental disorder in any one family was considered, all abnormalities were counted together, no matter how heterogeneous their nature in the light of modern knowledge. Under these circum­stances it was not possible to get any clear understanding of the hereditary factor in psychiatric illness, nor was it possible for studies in heredity to give any substantial aid to clinical psychiatry.

   The great work of Kraepelin laid a firm foundation for genetical work, and its solidity was shown by the fact that his clinical syndromes were found to be fairly pure genetically. In the families of schizophrenics only schizophrenics were found in significant excess, and the same was true of other syndromes. Nevertheless work in heredity suffered from the tendency to look for information from the accumulation of pedigrees, and to expect to discover modes of inheritance which followed classical Mendelian rules. The preconception of character­inheritance still interfered with a grasp of the fundamental mechanisms. The one mental disease whose hereditary basis was then clearly understood, Huntington's chorea, seemed to be a model which all others in time might be found to follow. In the early studies made by the Rüdin school in Munich attempts were repeatedly made to fit empirical findings with some hard and fast scheme of inheritance, and the statistician Weinberg was set to building up complex theories, such as that schizo­phrenia depended on two recessive and one dominant gene, which nowadays look absurdly naïve.

   It was the achievement of the Rüdin school to carry out a series of investigations which depended on no theoretical assumptions of a theoretical kind. A number of studies of the frequency of the various mental disorders in the general popu­lation established norms against which pathological material could be compared. This basic work has not been matched n any other country. The task was also taken up of finding the frequency of the several distinct psychoses among the relatives of various kinds of homogeneous groups of propositi, i.e., the frequency of schizophrenia among the sibs of schizophrenics, etc. It was a disappointment that these empirical figures did not fit Mendelian ratios, but a great practical advance was made, as a hereditary prognosis could be given, which would be of ser­vice in advising the relatives of patients about the wisdom of marriage and procreation. The necessary statistical foundation was also laid in developing the techniques of evaluating popu­lations at risk in groups of persons of very various age.

   While genetical work was still in this very rudimentary stage, and it was hardly possible to say more than that there was a hereditary basis for the endogenous psychoses, a fanatical philosophy led to a perverted eugenics, and the introduction in Germany of mass compulsory sterilisation. The cruelty and use­lessness of these measures have done much to discredit eugenics in general and to discourage genetical work in psychiatry all over the world.

   Nevertheless, the start that had been made led to further extensions, and to the introduction of one very significant new technique. Work proceeded on such vital problems as the fer­tility of the mentally abnormal, the heterogeneity of schizo­phrenia, the progeny of crosses between homogeneous and heterogeneous psychotic strains, the existence of assortative mating. Attempts to obtain useful results from the extension of the empirical method into familial relations of psychopaths and neurotics largely failed. It was left to the new technique, that of twin study, to open up the ground in this difficult field.

   Twins had long been of medical interest, but relatively small practical use of them had been made until a valid method of diagnosis of ovularity, independent of such obstetrical methods as examination of the placenta, had been found. Siemens laid the foundation for diagnosis by anthropometry, by the com­parison of bodily measurements, eye‑colour, etc., and von Verschiler showed that these methods were in fact more reliable than the older obstetrical ones. Rapid exploitation of this tech­nique followed; and a great range of valuable work has been done by such men as Lange, Luxenburger, Kranz, and Konrad in Germany, and in the U.S.A. by Rosanoft and Kallmann. Among the psychoses it was found that two‑thirds to four‑fifths of all uniovular twins of schizophrenics and epileptics were respectively either schizophrenic or epileptic; and the investi­gations of the twins of criminals showed that the hereditary factors were predominant in character‑formation, though less important than was at first supposed in social behaviour. The significance of heredity in the development of personality, well established as it is, has still not received adequate appreciation in general psychiatry, where this concept has run counter to psychodynamic trends derived from the immense and increasing popularity of psychoanalysis.

   In the field of mental deficiency it was left to British rather than German investigators to make the significant advances. The medical approach to the problems of mental deficiency was inadequate to do more than to establish the existence of a specific genetic factor in such rare conditions as amaurotic idiocy, epiloia and phenylketonuria. The sharp distinction which must be drawn on etiological grounds between pathological and physio­logical forms of mental defect, between the idiots and imbeciles and the feeble‑minded, appeared only from quantitative studies on the firm foundation of an adequate statistical approach developed by the Pearson school of biometricians. Much more was gained from the study of normal populations than from familial or even twin studies of the abnormal. The existence of precise, reliable and standardised intelligence tests, which we owe not to medical men but to psychologists, provided the basis for theories of multifactorial inheritance, which had been beyond the mental horizon of the empiricists.

   The appreciation of the significance of multifactorial inherit­ance in explaining the hereditary background for conditions of a quantitatively variable kind which are found widely distri­buted in general populations, has altered our ideas about the genetics of bodily constitution, temperamental traits of person­ality and such abnormalities as psychopathic personality and neurosis. Genetics here has come to the help of clinical psychiatry, and has taught us not to expect to find cleanly de­limited types of bodily build such as pyknics, athletics and asthenics, but rather dimensions of variation; and the same concept may be extended to cover such quantitatively variable features as personality traits and resistance to physiogenic and psychogenic stress. The approach that can be seen between the sciences of genetics, psychology and psychiatry, and the exist­ence of precise and flexible statistical weapons, indicate wide horizons for future work.


   Genetics is one of the basic biological sciences, and progress in psychiatric genetics is a contribution towards the building of a sound biological foundation for psychiatry. We must expect from genetics rather an elucidation of the nature of our problems than an immediate aid in prophylaxis and treatment. This does not mean that some application of theory cannot already be made. Kaliman [1] has recently made the very interest­ing suggestion that twin work in schizophrenia might be the starting point for fundamental work on the pathogenesis of the disease. Any pair of uniovular twins that can be discovered, of whom one only suffers from schizophrenia, should be subjected to thorough investigation, e.g., along biochemical lines, in the expectation that differences between the healthy and the diseased twin might throw light on the nature of the disease process. Such an experimental procedure might well be much more pro­ductive than mass studies on large numbers of patients, not specially chosen for clinical uniformity.

   The application of theory to eugenics and the prophylaxis of mental disorder has recently been discussed by Fraser Roberts and Slater.[2] In the psychoses and specific forms of mental defect such application is best confined to advice in the individual case; a wide social application is only possible and desirable in physiological mental defect. There are good grounds for believing that there is a negative correlation between intelli­gence and fertility, that, in fact, present tendencies are highly dysgenic. The genes responsible for multifactorial inheritance are very susceptible to quite minor selective processes, such as are involved in social traditions and public custom; and changes in the genetical constitution of populations, rapid on the his­torical scale, are quite possible. An attempt to convert a pre­vailing dysgenic into a eugenic tendency would have to be based, not on such measures as the sterilisation of defectives, which would only affect a fringe of the population, but on an alteration of the general attitude towards reproduction through­out the public.

   In mental deficiency a relatively small part of the total social disability is caused by the organic disorders responsible for idiocy and imbecility, the really important part being caused by non‑pathological feeblemindedness and backwardness. In the same way in general psychiatry it is the neuroses and psycho­pathies rather than the psychoses which produce the major amount of social incapacity, ill‑health, unemployment, dis­ordered marital and parental relations, unhappiness and friction. The neuroses and psychopathies are no doubt in part environ­mentally determined; in so far as they are genetically caused, the responsible genes must be multifactorial. Genetical research, therefore, should be based in the first place on a study of normal variation. Little advance could be made with intelligence until estimates could be made of normal means and standard devia­tions; analogous estimates are needed for a solution of the problems of temperament and personality. It seems likely that the best approach would be along physiological lines.

   It is to be hoped that one day, by the investigation of random samples of the population, we shall have reliable means and standard deviations of the principal physiological variables, and their inter‑correlations, and their responsiveness to various types of stress. One thinks of such things as blood pressure, adrenalin concentration in the blood, electrical skin resistance, E.E.G. wave pattern, speed of autonomic and neuro‑muscular reaction, etc. We would then have a foundation for the determination of such neurotic dispositions as the anxious temperament. Family investigations would then give us an idea of how and how far these qualities were inherited. The prospect of arriving at such standards seems a distant one, as few investigators are interested in the study of the normal, and fewer still in the statistical investigation of populations. Although it is a less fundamental approach, we are likely to get help sooner from psychology than from physiology. Psychologists are accustomed to thinking along these lines, and it would seem natural to them to take methods of investigation which have proved so successful in the study of intelligence for use in the study of temperament. A valuable start in this direction has already been made, notably by Eysenck; [3] the application of his methods to genetically related individuals, such as pairs of sibs, seems a promising line of attack.

   The studies in fertility which have proved of such social sig­nificance in the matter of intelligence might also be extended into the psychiatric field. So far the only work here has been on the psychoses. We have no knowledge about the fertility of neurotics and psychopaths; and the study of clinically differ­entiated groups might prove very important. It may well be that there are changes going on, not only in national intelligence but also in the national character.

   The significance of electroencephalographic work for the physiology of personality has been mentioned; genetical aspects will also prove important. From the work of Lennox and the Gibbses [4] the general form of wave pattern is largely determined by heredity; the records of uniovular twins are so alike that the ovularity of a twin pair can be diagnosed with only a small frequency of error. The working out of normal standards both in resting record and under response to stress is needed, and from the genetical point of view would be best in the quanti­tative form that the advent of the frequency analyser promises. From normal standards one might then proceed to the investi­gation of familial relationships and to the study of the abnormal. It does not seem likely that we shall get far on the present basis of a purely qualitatively estimated "non‑specific abnormality," and "dysrhythmia" is likely to remain an enigma so long as such arbitrary concepts are used. The association of E.E.G. abnormalities with psychiatric conditions other than epilepsy, which have a hereditary predisposition (such as obsessional states), suggests that constitutional variation in E.E.G. patterns will have many points of contact with clinical psychiatry.

   The E.E.G. is but one of the new techniques which may have a profound effect on the advance of psychiatric genetics; serology is now in the position to add another, of which perhaps even greater things can be expected. There are now about eight chromosomes which have markers in the form of genes of known and easily identifiable effect, and of sufficiently frequent appearance. The sex chromosome is readily identified, and there are satisfactory markers for other chromosomes in the blood factors, ABO, MN, Rh, P, and in the gene controlling the ability to taste phenyithiocarbamide as a bitter. If we were to follow the inheritance of a psychiatric abnormality such as Huntington's chorea, which depends on a single gene with a high rate of manifestation, and to couple with it tests for the presence of these known markers in affected and unaffected members of the family, we would therefore have one chance in three of finding that the psychiatrically morbid gene lay on the same chomosome as one or other of the markers. Refined statis­tical methods have been developed by Haldane, Penrose and others, which allow the efficient use of family material for demonstrating the presence or absence of linkage. Some use has already been made of these methods, though not in psychiatry, and there is evidence that the gene for sickle‑celled anamia is on the same chromosome as the MN genes. We have therefore taken a first step in the building up of a chromosome map for the human autosomes. Thanks principally to Haldane, the relative positions of a number of genes on the X chromosome are already known.

   If a morbid gene can be linked with a marker the possibilities of future work of greater precision are much enlarged, and such work might have the most radical consequences for psychiatry. In the realm of practical eugenics it would be of great signifi­cance. It is at present impossible to detect the carriers of the Huntington gene before they develop the disease, though the work of Minski and Guttmann5 supplies the hope that some­thing might yet be done in this direction along clinical lines. If, however, the Huntington gene was found to be closely linked with an easily detectable marker, it would become possible in Huntington families to predict with some accuracy which mem­bers were susceptible to the disease, and to warn them against reproduction.

   When we turn to the perennial problem of schizophrenia such practical possibilities are dwarfed by the theoretical poten­tialities. It is still much in dispute whether there are many different genes responsible for schizophrenic psychoses, or per­haps only one, and whether recessivity or dominance is the rule. Linkage investigations extended over a very large number of families in which more than one member was schizophrenic might well show decisively whether we must think of the genetical basis of the disease as homogeneous or heterogeneous; and in the latter case linkage studies would give us a firm foundation for the description of tiologically distinct syn­dromes. itiology once established, we could then proceed along the lines of biochemical investigation, and, by other appropriate methods, towards a solution of the pathogenesis.

   For many years, on the original initiative of Sir Frederick Mott, there has been accumulated at the Central Pathological Laboratory at the Maudsley Hospital (now the Research Labora­tory of the Institute of Psychiatry of the University of London) a card index of all patients admitted to the hospitals of the L.C.C, Mental Health Service (now hospitals of the various Metropolitan Regions) who were known to have had a relative similarly admitted at one time to one of these same hospitals. The treasury of material has never yet been properly exploited. One of the most profitable lines would be to use it for the link­age experiments already discussed, combining that with a clini­cal evaluation of the syndromes shown by genetically related individuals.

   Work of this kind was done by Kraulis [6] in Finland, although he made no use of linkage techniques, and his material is far from fully described. Nevertheless, he showed that the classical distinctions between manic‑depressive and schizophrenic ill­nesses did not always maintain themselves in individual families, and that related individuals did not always show the same type of psychiatric picture. Work along these lines would have the valuable end of showing what is genetically constant in particu­lar syndromes, and what features must be put down to other causes than the specific morbid genes.

   Many other uses might be devised for the Mott material, such as the correlation of ages of onset between sib pairs, a technique which in the hands of Haldane yielded information of theo­retical significance in Friedreich's ataxia; or an enquiry into the influence of age in affecting the form of symptomatology of mental illnesses. Valuable for genetical purposes would be an enquiry into the children of parents both of whom at some time became psychotic; this would, of course, involve extensive field work. It is not yet known how such accumulations of data as the Mott material will be affected by the regional organisation of the National Health Service; it is by no means unlikely that the facilities for making and for using such collections will be increased. In the University Institute of Human Genetics in Copenhagen, under the direction of Professor Tage Kemp, it has been possible to collect primary notifications of all cases of hereditary diseases occurring in Denmark, so that the material is gathered for comprehensive and nation‑wide studies of individual conditions. A large organisation is involved, with intimate co‑operation with the various hospitals of the kingdom. It may not be possible to establish anything on so large a scale in this country. Nevertheless, the nuclear collection, which was so far‑sightedly laid down and fostered in the one‑time Central Pathological Laboratory, may prove a seed from which a profit­able harvest may perhaps one day be gathered.



[1] Kallmann, F. j. (1948), J. Ment. Sci., 94, 250.

[2] Roberts, J. A. Fraser, & Slater, E. (1948), Eugen. Rev., 40, 62.

[3] Eysenck, H. J. (1947), Dimensions of Personality. London.

[4] Lennox, W. C., Gibbs, E. L., & Gibbs, F. A. (1942), Arch. Neurol. Psychiat., 47, 702.

[5] Minski, L., & Gutmann, E. (1938), J. Ment. Sci., 84, 21.

[6] Kraulis, W. (1939), Allg. Z. Psychiat., 113, 32.