Trends in Psychiatric Genetics in England 

In Expanding Goals of Genetics in Psychiatry (New York, Grune and Stratton, 1963)


SINCE EARLY IN 1960, reports of original observations on human chromo­somes have been appearing in the British medical journal, The Lancet, at an average rate of nearly one a week. There have, in fact, been so many new studies of this kind that a number of the reports, although presenting significant scientific information, have been confined to the span of a letter to the editor. In addition, there have been reviews, annotations, letters of comment or criticism, similar publications in other British journals and, of course, two important specialized symposia. [5,18]

   Since Britain is only one of a number of countries where chromosomal investigations have been pursued, the growth of this particular aspect of medical genetics may at the present time be regarded as almost explosive in its rapidity. Not only is it the case that established centers of research into human genetics have turned over part of their resources to work of this kind, but we also find that men whose background has not been in genetics proper ‑ for instance, clinical pathologists in some of our great hospitals have turned to cyto­genetics as a field of extraordinary promise. While these developments have brought genetics into the center of the picture in the broad field of general medicine, they have also done much to bring psychiatry into closer view for many workers in the general field. Actually, there is not one of the chromo­somal anomalies so far discovered which is not likely to be associated with psychiatric abnormality.

   The effects have not been less impressive in the psychiatric world itself. Mental deficiency, so long an area of relative neglect, now assumes a role of central importance. It is here rather than in the pathological processes of later life, that the biggest contribution to our understanding of etiology is likely to be made.

   It is as yet quite impossible to say how far and how fast further develop­ments along this line will take us. However, we may confidently look forward to a rapid accumulation of items of new information, which, bit by bit, will sum up to constitute a map of the territory we have now begun to explore. We have already seen the adumbration of hitherto unrecognized syndromes, associating anomalies of particular chromosomes with recognizable clinical patterns, to add to the one certainly known case of Down's syndrome. We shall be able to estimate their individual frequencies, to begin with, in such special classes of the population as the mentally defective or the infertile, and finally in the general population as a whole. Rapid improvements in technology will presumably enable us to implicate more and more chromosomes in the line of advance. There is definite hope, too, that regularities will be discovered in the mode of expression of the anomalies of each of them, in the frequencies of chromosomal mutations, in the relation of these frequencies to maternal and paternal age, and so forth. Finally, it may be expected that work of this kind will lead to the beginnings of an understanding of the mode of development of the pathological processes set in motion by the abnormal chromosomal constitution.

   Two pathways can be thought of as appropriate and promising. One might, for instance, adopt one of the techniques of animal genetics. A systematic study of early human embryos might end in the discovery that specified chromosomal abnormalities resulted in a perversion of the normal processes of development taking place at a particular stage in the life of the embryo and affecting particular tissues. In this manner, the chain of cause and effect might be followed through all its ramifications. Grüneberg's brilliant elucidation of the developmental process in the case of the lethal cartilage anomaly in the rat provides a master‑example of this approach. [9]

   Another pathway would be to apply to chromosomal anomalies the techniques of enzyme chemistry. In our own unit, some work of this kind has been done. In a study of the mothers of patients with Down's syndrome, Coppen and Cowie [3] were unable to confirm the observation by Ek [7] that there were indications of thyroid abnormality in a raised level of protein‑bound iodine in the serum. They did find, however, that women who had borne a mongoloid child at the age of 27 years or younger distinguished themselves from the older mothers by a raised androgyny score. An increase in the androgyny score was also shown by mothers who had had multiple miscarriages.

   This observation led to investigation [15] of steroid excretion in the mothers of mongoloid children, which confirmed the existence of a chemical difference between mothers with earlier and with later maternities of this kind. Mothers who had borne a mongoloid child early in life showed a remarkable increase in the ratio of output of dehydroepiandrosterone to total output of 17-ketosteroids. The mean level was about twice that of the other mothers, and there was no overlap.

    There are a number of possible explanations for this finding. One of them would be that the older mothers are constitutionally normal, while many of the younger mothers may be carriers of a balanced translocation, a chromosomal deviation from normality insufficient to cause any pathological appearance, but sufficient to lead to a difference in steroid metabolism. Hamerton and co‑work­ers [10] have found that an increased incidence of mongolism in the sibships of mongols is practically confined to the families of the younger mothers.

In this brilliant future, are there any considerations which should act as a warning or as a caution? It would certainly be disastrous if all effort in human genetics were confined to chromosomal studies. In such well‑investigated species as the fruit‑fly and the mouse, it is known that of all observed variance which can be put down to hereditary constitution, only a small fraction can be allocated to gross chromosomal deviations. By far the larger part of the observed variation is associated with microgenetics, with single gene effects, the effects of combinations of genes, position effects, genetic‑environmental interactions, and so forth. In man, too, we can estimate that at the end of the story a similar relationship will hold. If expansion in chromosomal genetics were to lead to the neglect of, let us say, biochemical genetics, we should be making a grave error.

   I doubt that such a mistake will be made. The main development which has taken place so far has led to the influx into genetics of workers, money and laboratory resources from outside the genetic field. Among workers within that field, no great diversion has so far occurred. Furthermore, the psychiatrist can point to the fact that people mainly want to do and actually do the kind of research work for which they are fitted by temperament and special gifts. For instance, mathematicians are not likely to start looking down microscopes. One of the great advantages of chromosomal work is that the techniques of cell culture and histology rely on the gifts with which humankind is best endowed, the ability to manipulate and the ability to see, and hardly at all on our more stunted capacities, such as that for abstract conceptual thinking. If men of a practical turn of mind are enabled to play a larger part in the development of our science, it will be greatly to our advantage. There will always be plenty of tasks for the others.

   These reflections might perhaps be interpreted as motivated by the need to console ourselves, for those of us who clearly are going to play no personal part in the exciting world which is opening up. If so, we surely have stronger grounds for hope. No major development occurs in science with its effects being shown in a single field. Work in frontier territories is always beneficially affected, and ideas and techniques which penetrate across the border are rapidly adopted. An illustration can be given from work in my own unit. [17]

   Many years ago, Theo Lang suggested that human homosexuality might represent, in part or in toto, a form of intersexuality. The idea was supported by the finding of an elevated sex ratio, with an excess of males, among the sibs of male homosexuals. Only in recent years has a more critical test of this hypothesis been provided by the technique of nuclear sexing, and the result was negative. Pare [13] reported that male homosexuals were chromatin‑negative, corresponding in their lack of Barr bodies with the normal male. However, this finding alone was not sufficient to dispose of the idea altogether that some anomaly at the chromosomal level may play a part in the causation of homo­sexuality.

   It is well established that there is a very marked shift in both birth order and maternal age in the case of Down's syndrome, and there are suggestions of heterogeneity in the increased variance of maternal age in Turner's and Klinefelter's syndromes. [14] Hence, an investigation of the given background factors seemed indicated in homosexuals, and such a study was made in a series of male homosexuals admitted to the Bethlem and Maudsley Hospitals between January 1, 1949, and December 31, 1960. [17]

   In this study, the maternal age associated with 355 out of the 401 homo­sexuals admitted was found to have a mean value of 31.3 years, as compared with 28.5 years for the general population, and 36.7 years for Penrose's material of patients with Down's syndrome. [14] In the homosexuals, the distribution of maternal ages was such that it might have been practically duplicated by a mixture of one mongoloid person to two members of the general population, and the variance of 50.8 was actually higher than that observed with mongoloid cases.

   Maternal age is not a very convenient statistic with which to work, since it is subject to a variety of influences affecting marriage and procreative patterns in the population. Position in birth order is much easier to study and is closely related to maternal age. With respect to birth order, adventitious influences such as affect the statistics of maternal age are compensated for, with the birth order of the sibs constituting control material. If an individual comes mth in order in a sibship of n individuals, then the ratio (m ‑ l)/(n ‑ 1) can readily be calculated, together with its mean, variance and standard error. The limits of the ratio are 0 and 1, and the expected mean is 0.5. In terms of this statistic, the birth order of 389 male homosexuals was 0.58, a figure which differed from the expected mean of 0.5 by more than three times its standard error. In other words, homosexuals tend to come late in birth order, and this finding suggests that there may be a certain proportion of bearers of chromo­somal anomalies among them.

   This investigation may have been rather limited in scope, but it illustrates the fructifying influence of advances along one line on research along other lines. It is also mentioned here for the purpose of drawing attention to a tool of research which might be more widely used by way of a sieve, to sort out syndromes worth special genetic attention.

   The knowledge which we may hope to gain from the study of chromosomal anomalies represents a breakthrough, by which psychiatric material will for the first time be used to provide the basis of a fundamental advance in human genetics. In psychiatric genetics, the debt is usually in the opposite direction, and the advances which are made are contributions to psychiatry rather than to genetics. Such nosology as there is in psychiatry depends as much on genetic work as on any other basic scientific discipline. It is the genetic findings in schizophrenia which provide the main justification for the present interest in the biochemistry of the psychoses.

   The immediate prospects in schizophrenia are not entirely encouraging. With a lot of work done, nothing very distinctive has been discovered so far. We are up against the difficulty that any anomalies are as likely to be secondary to the total disturbance of normal life caused by the psychosis, as primary to the psychotic process itself. The outlook is probably better with the affective disor­ders, where investigations of water and electrolyte chemistry offer hopes of getting to an understanding of the basic pathology. When the time comes for the chemists to lay their hands on a good working hypothesis, we must hope­fully look forward to the acuity of clinical research workers who will try it out on genetically defined groups. If there is a definable anomaly in the chemistry of the cyclothymic, it becomes of the highest importance to see whether the same or related anomaly is to be found in his sibs.

   The vexing question of the heterogeneity of schizophrenia remains a live issue. Personally, I belong to the school which regards the bulk of schizo­phrenics as being related to a single genetic and pathological syndrome, This hypothesis would not exclude the existence of a twilight area of heterogeneity surrounding the nuclear group. In clinical work in a neurological hospital, I have been greatly struck by the occurrence in long‑standing epileptics of paranoid psychoses very closely resembling paranoid schizophrenia.

   With my colleagues at the National Hospital,[2] we investigated the histories of 69 such patients. It was found that the bulk of them were temporal lobe epileptics, with some degree of cortical atrophy being a very common feature. There were, however, some clinical symptoms which seemed to differ from those of the common schizophrenic picture, especially the relatively satisfactory preservation of capacities of affective response. On follow‑up, the course of illness also tended to differ from that seen in schizophrenia. Deterioration of personality occurred with great frequency, but generally towards an organic type of personality defect, rather than a schizophrenic one.

   The genetic findings in this study were even more distinctive. While the incidence of schizophrenia was not increased in the parents and sibs of "schizophrenia‑like" epileptics, the incidence of epilepsy in their families ap­proximated the rate generally observed in the families of what used to be called "symptomatic" epileptics, Apparently, there is considerable justification for the view reached by Gruhle[7] many years ago. According to this view, it may be assumed that the psychotic illness of epileptics, despite its close resemblance to schizophrenia, is not schizophrenic at all.

   A different conclusion was reached by Kay and Roth,[1] who recently studied 99 paraphrenic patients over the age of 60 (42 patients from a British hospital and 57 patients from Stockholm) and compared these probands with matched cases of affective and organic illness. Among social and environmental factors which were reported as having been prominent in causation, deafness and social isolation were found to be particularly common. By contrast, evidence of a genetic predisposition seemed correspondingly attenuated. For instance, there was only a 3.4 per cent incidence of schizophrenia among the sibs and children of the 57 Stockholm probands.

   Although this figure might have been due to heterogeneity of the clinical material, the investigators rejected this explanation. They claimed that all the cardinal symptoms of schizophrenia were shown by their cases, and the indica­tions of hereditary loading were thought to have been more evenly spread than, say, in Stenstedt's study of involutional depression. [20]

   In line with this pattern of reasoning, Kay and Roth regarded their genetic findings as most easily explained on a multifactorial hypothesis. In the pre­disposed personalities of paraphrcnics, they say, we are probably dealing with extreme variants of the normal rather than with a rare and specific constitution. Monohybrid theories have to be buttressed by ad hoc hypotheses of reduced manifestation, influence of polvgenes on age of onset, etc., and no theory of this kind is entirely satisfactory to the investigators. Instead, they propose that in the light of the new evidence which is now being accumulated about the importance of environmental factors, the simpler hypothesis that the predisposi­tion to schizophrenia may be a graded character, depending on quantitative variation, merits more consideration than it has yet received. The all‑or‑none phenomenon, by which the predisposed individual is transmuted into someone recognizably suffering from schizophrenia, could be explained by threshold effects. Transitional stages between the schizophrenic and the normal are to be found in the formes frustes, paranoid psychopathy, and so forth. Conditions known to be due to single genes are probably less influenced by the environ­ment than schizophrenia appears to be. The social circumstances, which have been found to be associated with schizophrenia in early life, are likely to prove to be causes of the illness, and not just effects, although their contribution may be less than it is in paraphrenia.

   Since this is an interesting line of thought, it would be possible to find further support for it. According to Edwards,[6] it appears that the familial relationships of conditions showing quasi‑continuous variation may closely resemble those seen in single‑factor inheritance with reduced penetrance. He writes: "If we consider the simple cases of multifactorial inheritance, and of an abrupt threshold such that a proportion p of the population lies beyond it, then the intensity of familial aggregation shows a simple approximate relation­ship to p" (page 65). With a correlation coefficient of 0.5, over a wide range of values of p, the incidence in first‑degree relatives approximates the square root of p. Edwards continues: "It is interesting to note that in the range of incidence between 0.1 per cent and 1.0 per cent, within which lie epilepsy, schizophrenia, diabetes, spina bifida, anencephaly, pyloric stenosis and mental deficiency of unknown cause, the incidence in first‑degree relations expected on a multifactorial hypothesis is very similar to that found" (page 67).

   In the proceedings of a special symposium of the kind we have the pleasure of attending here, one cannot possibly think of genetic contributions to psy­chiatry without touching on the vast field of twin research. None of us questions the value of the contributions which have been made through twin studies in the past. We are gathered at this institute to honor the work carried on by Dr. Kallmann and his school and to acknowledge the increasingly general recognition among psychiatrists that genetic work, and particularly work with twins, has done much to clarify fundamental psychiatric problems. Of course, we may ask the question whether there should be the same level of scientific investment in work on twins in the future as in the past. My own opinion is that the work should go on, but should be extended to cover new territory, rather than go over old ground.

   In our twin research unit at the Maudsley Hospital, one of the two main groups of twins presently investigated by my colleague James Shields [10] is a series of twins from the general population, ascertained through a television program. This study was organized in such a manner that twins, who were willing to subject themselves to inquiry as volunteers, sent their names to the British Broadcasting Corporation. Among them were 42 members of MZ pairs who upon separation in early life had been brought up apart. Matched with an equal number of nonseparated MZ twins who served as controls, the given pairs were further compared with a series of 28 DZ pairs of the same sex, who had been separated early in life. On the whole, the design of the study has been such as to show up those environmentally determined differences in personality and life story which might be rooted in early childhood life. Despite some degree of repetition, this study is distinguished from the well‑known investigations of Newman, Freeman and Holzinger [12] in that the number of cases is larger, comparable controls are available, and the line of investigation has been rather different.

   In general, our samples showed almost equal similarity between members of pairs in both the \lZ groups, whether separated or control, while members of the DZ pairs differed from one another to a much greater extent. This finding holds for all data on which correlation coefficients could be calculated, includ­ing height, tests of verbal and non‑verbal intelligence, and tests of extraversion and neuroticism on Eysncks Maudsley Personality Inventory. The only, excep­tion was the correlation coefficient in respect to body weight in females, which was much smaller in the separated MZ pairs than in either of the other groups.

   With a more subjective rating of personality, environmental effects were more clearly discernible. As a rule, the MZ control twins were found to be more alike, though not significantly so, than the separated pairs. An attempt to identify the nature of this environmental factor led to the discarding of a number of plausible possibilities, but did provide a suggestion that the kind of home made some difference. The twin brought up in the psychologically less favorable home tended to have the poorer mental health rating.

   Both separated and control pairs in the MZ category disclosed substantial evidence of similarity in mannerisms, voice, temperament, tastes and other aspects of personalit, including sexual behavior. Personalities verging on the extreme with respect to quick temper, anxiety, emotional lability, rigidity and cyclothvmic tendencies were no less concordant in the one group (separated) than in the other (controls). The variables which produced personality dif­ferences ‑ some of which were quite noteworthy ‑ were usually found to be multiple. Since physical and social influences arising at different times in life seemed to interact with one another, the effect of the original home background as a differentiating factor tended to be obscured. In some pairs, twin partners remained obstinately alike, in spite of relatively large environmental differences.

   The other main series of twins on whom we have recently been working consists of nearly 200 pairs, in whom the twin index case had been given a diagnosis of neurosis or psychopathy by the admission service of the Bethlem­I\Jaudslev Hospital. A few months ago, a detailed report [18] was presented on those cases of the series diagnosed as "hysteria."

   Our main finding may be seen in the inability to confirm the status of hysteria as a genetic syndrome. At least, the data were of a very different order from what one would have expected if the disorder had been not hysteria but, let us say, schizophrenia. AS you know, twin studies have been criticized on the ground that high concordances must appear every time, whatever the group investigated, and hence ha e little meaning. Obviously, this view is quite unrealistic.

   In our series of neurotic twins, data are presently available on 52 twin index cases (from 39 pairs), admitted to the Children's Department of the Maudsley Hospital between January, 1, 1949, and December 12, 1958. Varying in age from two to 1 5 years when first seen, all twins have been followed up for a period of two to 11 years, except for four patients whom we have not yet been able to contact. It is unfortunate, however, that the given observation period has not been long enough to take more than a few cases beyond adolescence into adult life.

   Some problems arise here in deciding the method of calculating concordance. In eight of the 25 MZ pairs and in two of the 27 DZ pairs, the twin partners were referred together. Clearly, therefore, their referrals were not independent, so that the ordinary, method of counting doubly, ascertained pairs twice would not be justifiable. The alternative method proposed by Allen [1] would seem to be preferable. In this manner, we have found a concordance in MZ twins of 61 per cent, and in DZ pairs of 29 per cent. Clinically, the children constitute a heterogeneous material. Diagnosis in child psychiatry is usually allowed little importance. However, a rough and ready classification indicates that in seven pairs the principal problem was associated with mental deficiency, and in eight with an organic syndrome. The other cases were behavior disorders; 19 of the aggressive type, and 18 of the nonaggressive type. Differences in concordance between MZ and DZ pairs were more striking in the first two categories than in the others.

   The previously mentioned study, of separated twins may, serve as an example of the use of genetic methods, designed not only to elucidate the existence of conformities which might be attributed to genetic causes, but even more to discover dissimilarities which might be confidently attributed to the environ­ment. This has always seemed a very promising line of investigation, and it is a pity that when it is followed, the results are so often negative. It would be most welcome if the proponents of environmental etiology in psychiatry could provide us with some other controlled method of bringing out the psychological and social causes of illness and maladaptation. Such methods as have been used in the epidemiological and ecological approaches do indeed turn up interesting data, but always of a kind which lack critical value. For instance, the finding that schizophrenia is more common in the poorest social classes than it is at higher levels is somehow suggestive of environmental causation. However, the suggested environmental factors remain totally obscure, and the participation of genetic factors has by no means been excluded.

   The difficulties encountered in family studies of this kind are demonstrated by the still unpublished results of another study of our unit, [4] which dealt with the psychiatric history of the children of psychotic patients. According to environmentalist views, these children should be more than normally liable to the neurotic disturbances of childhood, and also to psychiatric disorder of any and every kind later in life. However, the data obtained have not confirmed this expectation. Whether one dealt in terms of historical events or in terms of psychological test results, the children of psychotic patients showed no greater tendency to neurotic disorders than did the children of a matched group of subjects from medical and surgical hospital patients. Nevertheless, the findings were not entirely, negative. Those neurotic disturbances which were observed seemed to occur preferentially during the two years following the onset of the psychosis in the parent.

   Our original plan had been to match this study with a completely different type of investigation. The subjects for investigation were to be girls admitted over a calendar year into a so‑called "classifying school," while the analysis was to be concerned with psychosis in the parents of patients with behavior disorder, rather than with neurosis in the children of psychotics. Unfortunately, however, administrative technicalities interfered with bringing this study to a satisfactory conclusion.

   In closing, let me say that this brief survey may not have portrayed a full picture of present trends in psychiatric genetics in our country. I do hope, however, that my report has given you a fair idea of what those workers in the field have been doing in recent years, with whom I have been in personal contact.



[1] ALLEN, G.: Comments on the analysis of twin samples. A. Ce. Me. Ge. 4: 143, 1955.

[2] BEARD, A. W., AND SLATEE, F.: The schizophrenic‑like psychoses of epilepsy. Proc. Roy. Soc. Med. 55: 311, 1962.

[3] COPPEN, A., AND Cowrn, V.: Maternal health and mongolism. Brit. M. J. 1: 1843, 1960.

[4] COWIE, V.: Children of psychotics. Acta psychiat, scandinav. 37: 37, 1961.

[5] DAVIDSON, W. M., AND SMITH, D. R. (Eds.): Human Chromosomal Abnormali­ties: Proceedings of a Conference Held at King's College Hospital Medical School. London, Staples Press, 1961.

[6] EDWARDS, J. H.: The simulation of mendelism. Acta genet. 10: 63, 1960.

[7] EK, J. I.: Thyroid function in mothers of mongoloid infants. Acta Paediat. 48: 33, 1959.

[8] GRUHLE, H. W.: Uber den Wahn bei Epilepsie. Ztschr. ges. Neurol. Psychiat. 154: 395, 1936.

[9] GRUNEBEEG, H.: Animal Genetics and Medicine. New York, Hoeber, 1947.

[10]  HAMERTON, J. L., et al.: Chromosome studies in detection of parents with high risk of second child with Down's syndrome. Lancet 2: 788, 1961.

[11] KAY, D. W. K., AND ROTH, M.: Environmental and hereditary factors in the sehizo­phrenias of old age ("late paraphrenia") and their bearing on the general problem of causation in schizophrenia. J. Ment. Sc. 107: 649, 1961.

[12] NEWMAN, H. H., FREEMAN, F. N., AND HOLZINGER, K. J.: Twins: A Study of Heredity and Environment. Chicago, Univ. Chicago Press, 1937.

[13] PARE, C. M. B.: Homosexuality and chromosomal sex. J. Psychosomat. Res. 1: 24?, 1956.

[14] PENROSE, L. S.: Parental age and non‑disjunction. In Davidson, W. NI., and Smith, D. R., Eds.: Human Chromosomal Abnormalities: Proceedings of a Confer­ence Held at King's College Hospital Medical School. London, Staples Press, 1961.

[15] RUNDLE, A., COPPEN, A., AND COWIE, V.: Steroid excretion in the mothers of mongols. Lancet 2: 846, 1961.

[16] SHIELDS, J.: Monozygotic Twins Brought Up Apart and Brought Up Together. Ox­ford, Oxford Univ. Press, in press.

[17] SLATER, E.: Birth order and maternal age of homosexuals. Lancet 1: 69, 1962.

[18] --‑: The Thirty‑Fifth Maudsley Lecture: "Hysteria 311." J. Ment. Sc. 107: 359, 1961.

[19] SMITH, D. R., AND DAVIDSON, W. M. (Eds.): Symposium on Nuclear Sex. London, Heinemann, 1958.

[20] STENSTEDT, A.: Involutional melancholia. Acta Psychiat. (suppl. 127),1959.