Eugen Bleuler’s Views on the Genetics of Schizophrenia in 1917

Abstract

In 1917, Eugen Bleuler published an article (Mendelismus bei Psychosen, speziell bei der Schizophrenie [Mendelism in the Psychoses, especially Schizophrenia]) in response to the recently published first systematic family study of dementia praecox (DP) by Ernst Rüdin, then working under Kraepelin in Munich. Although briefly commented upon by David Rosenthal in 1978, this article has never been thoroughly reviewed or translated. Of the many themes addressed, four are especially noteworthy. First, Bleuler argues that understanding the transmission patterns of schizophrenia in families requires definitive knowledge about the boundaries of the phenotype which he argues are unknown. Rüdin’s choice—Kraepelin’s concept of DP—is, he asserts, too narrow. Clarifying the genetics of schizophrenia is inextricably bound up with the problem of defining the phenotype. Second, Bleuler argues for the importance of “erbschizose” (literally “inherited schizoidia”) wondering whether his “4 As” or other “brain-anatomical, chemical, [or] neurological characteristics” might underlie the genetic transmission of schizophrenia. Third, Bleuler was deeply interested in the nature of the onset of schizophrenia, suggesting that environmental adversity could provoke “latent illness to become manifest.” It was important, he argued, to identify such risk factors and incorporate them into genetic models. Fourth, although not optimistic that current knowledge would permit a resolution of the transmission model for schizophrenia, he finds single-locus models implausible and at several points wonders whether polygenic models might better apply. A complete translation of the article is provided.

Second only to Emil Kraepelin (1856–1926), our current concept of schizophrenia has been most influenced by Eugen Bleuler (1857–1939). In Section IX on “The Causes of the Disease” in his 1911 opus Dementia Praecox or the Group of Schizophrenias,^1,2Bleuler discusses briefly a number of features of the possible role of heredity in the etiology of schizophrenia. However, 6 years later, in a dense 21-page article in the Swiss Archive of Neurology and Psychiatry entitled “Mendelismus bei Psychosen, speziell bei der Schizophrenie (Mendelism in the Psychoses, especially Schizophrenia),” ³ Bleuler describes in detail his views on and questions about the nature of the genetic risk for schizophrenia. I am unaware of any English translation of this article, and a bibliographic search found 31 published references to this report mostly in German or English, a sampling of which I here cite.^4–7 No article was found that presented, in any detail, the content of this article except a brief one and a half-page note by David Rosenthal in 1978 which commented on its likely historical importance.⁸ This essay seeks to make Bleuler’s essay available to contemporary scholars. I here summarize the views expressed on the genetics of schizophrenia by the man who named and carefully described the syndrome that remains at the heart of the disciplines of both psychiatry and psychiatric genetics.

Bleuler’s essay is challenging to summarize as it is not highly linear in an organization and the original German is sometimes obscure. Some of its richness is in its “side-comments.” I have imposed an artificial structure of sections to aid the reader and had to be selective in the material covered. A complete English translation is available in the supplementary appendix. I here use the terms dementia praecox (DP) and schizophrenia interchangeably.

The stimulus for Bleuler’s essay was the publication a year earlier of the first systematic study of risk of illness in the siblings of DP probands by Ernst Rüdin,^9,10 who worked in Kraepelin’s Department in Munich, Germany. Here is a succinct summary of Rüdin’s study:

The first major family study of schizophrenia, reported by Ernst Rüdin in 1916, examined 2,732 siblings of 755 probands, diagnosed according to the teachings of Kraepelin. This study, the goal of which was to see whether the segregation pattern of schizophrenia in siblings conformed to simple Mendelian expectations, was the first in psychiatry to use systematic ascertainment, proband correction and calculation of an age corrected risk of illness-the morbid risk (MR). The MR for narrowly and broadly defined schizophrenia in this sample can be calculated to equal 5.4 and 7.7%. “Other psychoses” – a heterogeneous category – were also common in these siblings (a MR of 5.1%) …. The risk for schizophrenia in siblings was significantly increased by a parental diagnosis of alcoholism, a history of schizophrenia in second- or third-degree relatives, and, particularly, by a parental diagnosis of “other psychoses” [see reference (¹⁰) (p. 338)]

Readers should be aware that Rüdin was an active Eugenicist throughout his adult life, and, 17 years after publishing this study, was funded by and collaborated with the German National Socialists from the time they came to power in 1933 until the end of the World War II.¹¹ The level of his moral culpability has been debated but it is undisputed that he advocated for and actively participated in the large-scale sterilization program for the “mentally unfit” undertaken in German in 1934.¹² For a summary of the genetics research program that he established in Munich, initially under Kraepelin’s leadership, see reference.¹³

Introduction

After a few introductory technical comments about Rüdin’s proband correction method, Bleuler dives into Rüdin’s main conclusion about the possible Mendelian transmission patterns of schizophrenia, initially focusing on the subsample of siblings where neither parent was affected:

Taking into account all these conditions, the siblings of schizophrenic patients with seemingly healthy parents yield an affliction rate [i.e., morbid risk (MR)] of 4.48% for schizophrenia and 4.12% for other psychoses. Based on this result, manifest schizophrenia cannot be a monohybrid [i.e. single-locus] Mendelian trait. The disease cannot be thought of as dominant, considering the rarity of cases of direct heredity. The “healthy” parents of schizophrenic children must therefore both be recessive heterozygotes, and 25% of the children would have to become ill, if manifest schizophrenia were a simple recessive trait. However, it could be possible that schizophrenia is a dihybrid [two-locus] trait, in other words, made up of two characteristics, which in the descendants of two-fold latent bearers of the trait occurs in the ideal ratio of 6.25% (1/16). That the observed rate of 4.48% is somewhat smaller than the requirement can easily be explained by the circumstance that most of the pronounced cases in every family group were not included and that the milder cases could escape the statistics. [see reference (³), (p. 21)].

Bleuler is here reiterating Rüdin’s conclusions, which are summarized in English in more detail in reference.¹⁰ Working from this two-locus recessive model, where all individuals would have 0 to 4 risk alleles, Bleuler suggests that while 4 would be required for classical schizophrenia, having 1 or 2 might result in “half-afflicted types,” which could

… be represented by schizophrenia-like and non-schizophrenia-like psychopaths [which in Bleuler’s time was much more broadly defined than today, referring to individuals with any kind of, typically non-psychotic, psychopathology], which we so often encounter in these families, and by means of anomalies, which give the impression as actual, but not schizophrenic psychoses, and according to Rüdin’s data, which in such families are almost as frequent as schizophrenia itself. [see reference (³), (p. 22)]

Bleuler is suggesting that a reduced genetic load in a two-locus recessive model could result in the nonschizophrenic psychotic cases and the disordered personalities seen in Rüdin’s sibling sample. He then reviews Rüdin’s data regarding the risk to offspring as a function of parental diagnoses:

Of children from a schizophrenic parent … they yield a ratio of 6.18% schizophrenic patients and 10.3% who are otherwise psychotic, totaling 16.47%…. If a parent is in some way mentally ill or alcoholic (but not schizophrenic), then 8.21% of the children are schizophrenic (instead of 4.48% in all families overall) …. If both the parents, or one of them, is mentally ill in some way, and if there is concurrent alcoholism in one or the other, then 14.81% of the children are schizophrenic, 7.40% psychopathic. If one of the parents is alcoholic and one is mentally ill, then the figures are 7.80% and 5.20%. If both parents are mentally ill in some way (including possible alcoholism), then 22.72% are schizophrenic, and no others psychotic in some other way …. Finally, if one parent is mentally ill in some way, and additionally one or both the parents are alcoholic, this yields the figures of 15.78% or 7.89%. [see reference (³), (p. 22)].

From these results, Bleuler draws the conclusion that DP

… does not only occur in connection with the origins of the particular illness in the family [i.e., DP], but also with other psychoses and with alcoholism, indeed, that the schizophrenic burden has hardly a closer relationship to the development of dementia praecox than the other psychopathic factors. [see reference (³), (pp. 22–23)]

Bleuler is questioning the specificity of the genetic risk for DP. He then quotes a seminal passage from Rüdin’s monograph that addresses a key issue for Bleuler—what is being inherited in these families?

It will be necessary to gradually find out, by means of many individual clinical examinations and numerous statistical groupings, which psychopathic and psychotic states are to some extent to be considered hereditarily equivalent and which are not. [see reference (³), (p. 23)].

So Rüdin suggests, with Bleuler’s concurrence, that further investigations of the essential nature of the schizophrenia phenotype need to be addressed from both clinical and statistical perspectives. Bleuler’s then raises another key issue: the nature of the etiologic pathways to schizophrenia. He writes:

If these studies are correct, then there is no single gene for schizophrenia, instead different things have to coincide to cause the illness to emerge. A polymorphous [e.g. polygenic] heredity can at this stage also not be excluded. [see reference (³), (p. 23)].

In this fertile passage, Bleuler posits that the emergence of full-blown schizophrenia requires exposure to more than one risk factor. Furthermore, he not only rejects a single gene model for the illness but also suggests that polygenic inheritance also should be considered.

Section 1

Bleuler’s introduction to the next section is disarming in its recognition of the depth of the ignorance that he felt about many key issues:

Nevertheless, we have not reached the goal, but have gained a sensible foundation, useful for further spheres, and now the questions can begin. However, we immediately encounter a confusion of difficulties … The most important questions are so closely related that none seem entirely resolvable without answering the others. [see reference (³) (p. 24)]

He decides upon a first line of attack, examining the problem of what we would now call “phenotypic definition.” He takes a position quite different from Rüdin’s:

The first basis of a heredity study should be the determination of the concept and the extent of the feature [e.g., phenotype] which here is the psychosis. But it is precisely these norms that are problematic in mental illnesses and, conversely, should be determined by heredity. Rüdin first demonstrated his method of working by means of the example of dementia praecox, because he believes that this psychosis can be determined with the greatest certainty. I am of the opposite opinion. I think that dementia praecox is a psychosis [the boundaries of] which is most difficult to determine. [see reference (³), (p. 24)]

Bleuler’s use of the term “norms” likely refers to discrete phenotypes then commonly subject to study by Mendelian methods such as Mendel’s height, color, and smooth vs. wrinkled seeds of the pea plant. Furthermore, when Bleuler was writing, several human traits had been proposed, with pedigree evidence, to be transmitted as Mendelian traits including ABO blood group, polydactyly, alkaptonuria, and Huntington’s disease.¹⁴ While Bleuler agreed with Rüdin that severe cases could be diagnosed with minimal difficulty, this does not apply to the milder cases which he thinks are more common:

… it is very easy to recognize in the qualitatively and quantitatively pronounced cases, most of which come to our insane asylums - and Rüdin selected these “purely based on the certainty of the diagnosis”. …. in the tentative beginnings, he wanted to narrowly define the concept of the illness being examined, so that the object of the investigation does not become “watered down.” However, in the case of dementia praecox, this caution is poorly applied, and will certainly lead to a false result. It is clear that dementia praecox is much more widespread than the cases in mental hospitals with an absolutely certain diagnosis in the sense of the current Munich School… one does not need to believe, like me, in an uncommonly frequent latent schizophrenia, and one does not need to accept as true that the cases not known to the mental hospitals are far more numerous than the ones that are known, to nevertheless be convinced of the incorrectness of such a restriction. It is a fact well-known to any psychiatrist who has had a few years of experience that doubtful diagnoses at the initial examination later reveal themselves to be schizophrenic to a very large extent. [see reference (³), (pp. 24–25)].

Bleuler suggests that it is fine, in a family study, to carefully select probands that are severely ill and diagnostically certain. However, examining family members, one will inevitably encounter diagnostically ambiguous cases which Bleuler feels will substantially outnumber the more classical cases. (Given the expansion of his concept of schizophrenia, Bleuler felt that many boundary cases should be considered schizophrenic, but Rüdin, committed to his mentor’s narrow concept of DP, classified them as “other psychoses.” In our summary of Rüdin’s study,¹⁰ we examine in detail these cases, 33% of whom Rüdin considered to have possible or probable DP.)

Bleuler then provides a “side-comment” about the causes of schizophrenia onset in a high-risk individual:

Poor environmental conditions, an intoxication, a crime, an agitation triggered by a coincidental situation and other incalculable conditions very often cause a schizophrenia, which had existed for a long time already, to come to light. If such a coincidence had not occurred, one would have known nothing of the illness. [see reference (³), (p. 25)].

Bleuler suggests that individuals with substantial liability to schizophrenia might, in a life devoid of major stressors, never manifest the full disorder. In genetics parlance, this would be “incomplete penetrance.” He returns to the problem of defining phenotypic boundaries: “Where we draw the line for schizophrenia-like psychopaths in the families of our patients, is entirely arbitrary” ³ (p. 25). He then questions the wisdom of selecting schizophrenia for Mendelian analysis:

If other plant and animal trait pairs [e.g., Mendel’s pea phenotypes] had not led us to Mendelian theory, we would never have thought of anything else in dementia praecox than what we now call intermediate heredity, i.e. heredity with infinite gradation in relation to the intensity of the illness. Thus, the facts do not give us the faintest indication for calling that, which by incidental events or somewhat greater intensity from a continuous ladder of qualitatively similar phenomena, stands out somewhat from the numerous other cases, a particular disease. [see reference (³), (pp. 25–26)]

Schizophrenia is not, Bleuler argues, a traditional genetic phenotype with clean boundaries. This is a core problem for genetic studies of schizophrenia:

Unfortunately, we cannot provide anything positive to replace the fiction: The boundaries of dementia praecox compared to health or compared to other psychoses and the neuroses are at this stage not yet known to us. We neither know the diagnostic threshold nor the individual symptoms or the picture as a whole. That is, we do not know how pronounced an anomaly in the sense of schizophrenia needs to be, to be utilizable as a certain sign of the illness. [see reference (³), (p. 26)]

This problem of defining phenotypic boundaries critically impacts on the results of statistical modeling:

If Rüdin had a wider concept of schizophrenia, he would arrive at very different results with regard to the important points: The ratios would not differ so much from the tentative provisional 25% of Mendel [for a recessive disorder]. In any case, the convergence could be considerable, as a fair part of other psychoses would be included … [see reference (³), (p. 27)].

Bleuler then uses Rüdin’s data on risk in offspring for different patterns of illnesses in parents to support his position for a broad phenotypic definition of schizophrenia:

Altogether, the familial relationships of schizophrenia with other psychoses and neuroses cannot be considered only from the point of view of di- or poly-hybridism. Precisely these with great probability indicate the correctness of an expanded schizophrenia concept. If a parent is not “schizophrenic”, but mentally ill in some other way, then there is not only an increase of the ratios of descendants who will be mentally ill in some other way … but also that of schizophrenia. In the sense of di-hybridism, this would imply that another mental illness or a split product of it, forms a component of dementia praecox. This observation could be explained by means of polymorphous heredity. [see reference (³), (p. 27)]

For genetic studies of schizophrenia, Bleuler strongly comes down on the side of the broad end of the broad vs narrow phenotype debate:

In any case, the “unclear cases” such as every psychiatrist encounters have a special significance. I believe that it is a lesser mistake, as far as dementia praecox is concerned, is to include all these illnesses than it is to exclude them. [see reference (³), (pp. 27–28)].

Bleuler then turns to the problem of etiologic heterogeneity—echoing the subtitle of his earlier monograph on The Group of Schizophrenias. In his 1917 article, he writes:

Initial insurmountable difficulties are provided by the delineation of the individual forms within the entire schizophrenia group. Here one conversely has to expect the definitive indication from heredity research alone. I am convinced that the concept of dementia praecox contains different illnesses [see reference (³), (p. 28)].

He suggests that sometimes mixtures of schizophrenia with manic-depressive insanity or with epilepsy run in families and perhaps constitute different illnesses. Later in the essay, he summarizes the concern that we “do not know whether the mental illnesses correspond to simple or compound hereditary traits” [see reference (³), (p. 40)].

Section 2

Bleuler then returns to a central problem:

The basis of all heredity research stands on shaky ground when it comes to schizophrenia: what is inherited? Is the clinical manifestation something that corresponds to the illness in the sense of a congenital anomaly? Is there something which is inherited which one can even call dementia praecox? These questions have not been addressed yet, not to mention answered. No one denies that among the conditions of the onset of the illness inherent predisposition plays an important role, although this has not been satisfactorily proven [see reference (³), (p. 29)].

He proposed two possible mechanisms. The first is more deterministic:

Is there thus a predisposition which only leads to a particular trait at a certain age? Such as those that lead to obesity or arteriosclerosis, or to dark pigmentation of hair or eyes, which increases under the most prevalent conditions during the course of childhood? [see reference (³), (p. 29)].

The other involves joint effects of genetic predisposition and other, likely environmental, stressors:

Or perhaps the illness develops with an existing predisposition and the concurrent addition of certain harmful effects? Is this finally the reaction of a pathologically predisposed mind to different harmful effects? What anomalies in the mental predisposition lead to this kind of reaction? [see reference (³), (p. 29)]

At this point, Bleuler introduces two important but difficult to translate terms. The first is erbschizose, which literally means “hereditary schizoidia” where schizoidia reflects our idea of “schizophrenia liability.” The second is sichtschizose which means literately “visible schizoidia.” I will use the original German, but Bleuler’s concept of sichtschizose is probably close to what we might call the “schizoidia phenotype.” Erbschizose is harder to translate into modern terms as Bleuler uses this more frequently to refer to what we might call an “endophenotype,” but elsewhere, his meaning seems closer to “schizoidia-genotype.”

He then asks what is the nature of the erbschizose? He begins by saying what it is not: “What I have called secondary symptoms [in his 1911 monograph] (content of hallucinations, of delusions …. and suchlike) cannot be inherited at all” [see reference (³), (p. 31)].

He then continues:

However, the inherited anomaly can be expressed in the primary symptoms [his “4 As”] …. Nevertheless, we do not know if this happens, as we do not understand these symptoms well enough yet. Most probably these primary symptoms do not correspond to the inherited anomaly itself …. The erbschizose thus still has to be found. [see reference (³), (p. 31)]

So, perhaps his famous “4 As” are the essence of the heritable liability to schizophrenia, but then perhaps not. He then ponders the possible relationship between erbschizose and sichtschizose. The erbschizose

… is connected with the sichtschizose by means of a long causal chain which is probably complicated by the involvement of many internal and external factors. It is no wonder that we cannot come to the hereditary unit by means of a straight path, if we start from the other end with the sichtschizose. [see reference (³), (p. 31)]

Selecting a definitive schizophrenia phenotype is so difficult, he claims, because of the length and complexity of the gene to phenotype pathway. What might the erbschizose be? It:

… does not need to be a direct mental anomaly. It can be a brain-anatomical, chemical, neurological characteristic, which manifests the illness as a symptom, be it on its own or with the help of internal or external influences ([see reference (³), (p. 31)]

In modern parlance, Bleuler suggests that the genetic abnormality in schizophrenia might be best conceptualized at several different scientific levels including the neurobiological and the mental/psychological. He then comes back to consider other factors contributing to schizophrenia onset:

Without a doubt, external influences can cause an already existing mental abnormality in the sense of dementia praecox to develop to a full illness, that is to cause a latent illness to become manifest[see reference (³), (pp. 31–32)]

He also suggests that internal factors may play a role:

Additionally, internal influences, accompanying mental predispositions, which in themselves have nothing to do with the gene of the illness, can possibly contribute. For example, a degree of sensitivity, which is distinctive of future schizophrenics, can reduce the adaptability of the patients, cause them to feel difficulties particularly strongly and contribute to the onset of the illness. [see reference (³), (p. 32)]

That is, psychological traits that are unrelated to the liability to schizophrenia could render vulnerable individuals more sensitive to the pathogenic effects of stressors. This could be a way in which the apparent genetic relationship between schizophrenia and alcoholism and other psychiatric disorders could be understood. That is, the liability to such disorders could contribute, in a nonspecific way, to the onset of schizophrenia by rendering individuals more stress vulnerable.

Section 3

I here present a selection of the ideas discussed by Bleuler in the final parts of his essay. He again turns to the importance of environmental influences unaccounted for in Rudin’s study:

The environmental influences can have an effect in the most diverse places. They can determine the illness, or merely manifest it. It is certain that they determine the onset of an attack and can determine the external picture. Conceivably there are environmental influences which prevent the onset of the illness. [see reference (³), (p. 32)]

He proposes a quantitative and potentially polygenic index of risk for schizophrenia:

The gradation of nuances of dementia praecox from healthy to its most severe manifestation can be due to different reasons. Firstly, many factors and their combination can simulate a continuous range. In particular, the severity of the disease can be co-determined by the other mental characteristics of their bearer, which can change in infinite numbers of ways…. It can also be conceived that several genes add up to a maximum of one appearance, while the presence of smaller numbers determine the same intermediate forms. [see reference (³), (pp. 33–34).

He suggests that the study of premorbid features as a way to understand better the nature of erbschizose:

To gain a closer understanding of erbschizose, we should study the characters of future schizophrenics. One could use good anamneses of people who are now afflicted with schizophrenia, additionally, the characters of children of schizophrenics and younger siblings, whose peculiarities indicate the later onset of the illness. [see reference (³), (p. 34)]

Finally, near the end of this essay, Bleuler provides a list of what would be needed to conduct a definitive Mendelian genetic study of schizophrenia. Three are noteworthy. In the first, he uses two terms–erbpsychose and sichtpsychose—that are parallel to terms noted above—erbschizose and sichtschizose, except they refer to the broader syndrome of psychosis. He notes that we would need “a well-defined clinical definition of dementia praecox, both in the clinical (Sichtpsychose) and in the genealogical sense (Erbpsychose)”³ (p. 35). To do this, we would need to know that schizophrenia had a “consistent clinical picture” or if it was “polymorphous,” we would need to know all the “equivalent clinical pictures.” We need to be able to correctly diagnose illness in affected cases. Second, we would need to understand “the causal relationships between erbpsychose and sichtpsychose.” That is, how do the inherited vulnerabilities for schizophrenia translate into the clinical syndromes? Third, “no environmental influence may cause a big change in the [morbidity risk figures], or it has to be qualitatively and quantitively discernible and be included in the calculations.” So, if large environmental risk factors are identified for schizophrenia, it will not be possible to clarify the genetics of schizophrenia without measuring the events and including them in a complete statistical model.

An appropriate conclusion to Bleuler’s essay is captured by this quote:

Nevertheless, it is not pleasant to see Rüdin’s high, as well as naïve, hopes dashed of imminently solving the heredity problem of psychosis. However, it is possible only to work fruitfully in this sphere if one is aware of the many possibilities and difficulties[see reference (³), (p. 36)].

Discussion

Major Points

Of the numerous themes touched on by Bleuler in this wide-ranging and prescient essay, five are noteworthy. First, Bleuler argues that to conduct a family study of schizophrenia with the goal of determining its pattern of genetic transmission, a clear definition of the phenotype is required. But, he suggests that the exact boundaries of the schizophrenia phenotype were then unknown, rendering any potential conclusions about the possible mode of transmission premature. He suggests, however, that Rüdin’s choice—Kraepelin’s concept of DP—was too narrow. He bases this opinion both on his own clinical experience and also on Rüdin’s data showing that the rates of schizophrenia were elevated in the offspring of parents suffering from a range of nonschizophrenic disorders. He concludes that an understanding of the genetics of schizophrenia is inextricably bound up with the problem of the definition of the syndrome. Progress in one cannot be made without progress in the other.

Second, Bleuler argued for the importance of clarifying what he called the “erbschizose,” which he seems to use both for our concepts of “genotype” and “endophenotype.” He wondered if his “4As” might qualify, but then suggests it might be a “brain-anatomical, chemical, [or] neurological characteristic.” He argued that identifying the nature of the “erbschizose” would greatly aid in clarifying the genetic transmission pattern for schizophrenia.

Third, Bleuler is skeptical about the etiologic homogeneity of schizophrenia—suggesting that it is likely several largely distinct disorders. Fourth, Bleuler was interested in the nature of the onset of schizophrenia. Environmental adversity could play a role, causing “latent illness to become manifest.” He not only considered it is important to identify such factors and incorporate them into genetic models. But he also noted that internal factors unrelated to schizophrenia risk could lower the threshold of onset. Indeed, this was one way in which it might be possible to understand the familial relationship between schizophrenia and other psychiatric disorders.

Fifth, although not optimistic that knowledge in his day would permit a resolution of the correct transmission model for schizophrenia, he noted that a single-locus model was implausible. At several points in the essay, Bleuler suggested that a polymorphous or polyhybrid model might apply. No details are provided for the specifics of such a “polygenic” model.

Historical Context

To put Bleuler’s comments in historical context, it is helpful to compare them with those of his contemporary, Emil Kraepelin. In his famous 1899 6th edition in which he first articulated his concept of DP,^15,16 Kraepelin commented on the frequency of “hereditary predisposition” in DP cases which he estimated is found in 70% of cases. In 1913, three years before Bleuler’s essay, in the third volume of his final 8th edition,^17,18 Kraepelin provided a few more details. A direct hereditary trait (ie, found in parents, grand-parents etc.) was seen in 34% of cases and he noted for the first time that DP itself is often familial but provides no figures. He noted, in agreement with Bleuler, the frequency of “striking personalities … eccentric individuals, failed people …” in the siblings of DP probands¹⁷ (p. 919). He also reviewed preliminary results from Rüdin’s family study, noting that Rüdin favored a recessive mode of inheritance and frequently found cases of alcoholism, manic-depressive insanity, and “eccentric personalities” in the siblings¹⁸ (p. 234). He made no other comments suggesting that Bleuler’s views about the genetics of schizophrenia were at a more sophisticated level than that of his famous contemporary Kraepelin.

Bleuler’s essay appears to have had a minimal impact on later research in the genetics of schizophrenia. In the 31 references to Bleuler’s article listed in Google Scholar (11/19), none included modern articles reporting empirical data or reviews on the genetics of schizophrenia. While this article is listed among Bleuler’s writings in Gaupp’s tributary review of Bleuler’s work published shortly after this death, the essay makes no mention of his review of Rudin’s study.¹⁹ Eugen Bleuler, himself, never appeared to be involved in empirical genetic studies of schizophrenia prior to or after this publication. This is surprising given the obvious attention he had paid to this problem at least in the years prior to 1917. However, his son—Manfred Bleuler—who also went into psychiatry and like his father obtained the position of the Director of the Burghölzli Hospital in Zurich—was a major figure in the genetics of schizophrenia publishing several major family and studies performed in both the United States⁴ and Switzerland.²⁰

This article should be reviewed with at least one caution in mind. In such endeavors, there is always a risk of writing “Whig History,” in which the author presents his topic as a story leading inevitably to the insights of the present day. I tried to counter this bias, but readers can best judge this by reading for themselves the entire article available online.