Clinical Review & Education

JAMA Psychiatry | Special Communication

From Many to One to Many—the Search

for Causes of Psychiatric Illness
Kenneth S. Kendler, MD

Supplemental content
The search for the causes of medical and psychiatric disorders has gone through 3 historical
phases. First, up until the mid-19th century, causes of illness were anecdotally recorded from
individual cases, resulting in long and diverse lists for all disorders. Second, in the latter half of
the 19th century, with the use of microbiological methods, single causes were found for many
infectious diseases that led to specific diagnostic tests, effective preventions, and, in some
cases, treatments. Causal thinking in medicine shifted from the earlier multicausal
approaches to monocausal theories of etiology. Indeed, proving monocausal etiology became
a way to establish the legitimacy of a disorder. Through the writings of Kahlbaum and Hecker,
psychiatry was deeply influenced by this monocausal perspective, the importance of which
was substantially amplified by a twist of fate: the increasing clinical importance of general
paresis of the insane throughout the 19th century and the eventual proof that it too was a
monocausal condition. However, in the mid-20th century, the third phase began. With
decreasing deaths from infectious diseases, epidemiology and clinical medicine shifted to a
chronic disease model in which paradigmatic disorders, such as cancer and cardiovascular
Author Affiliations: Virginia Institute
disease, were shown to be highly multicausal. Biostatistics evolved from deterministic to for Psychiatric and Behavioral
probabilistic models of disease risk factors. Paradoxically, at this time, biological psychiatry, Genetics, Department of Psychiatry,
then rising to dominance in American psychiatry, vigorously pursued monocausal theories, Virginia Commonwealth University,
first of neurochemical origin and then of genetic origin. We were trying to establish the Richmond; Department of Psychiatry,
Virginia Commonwealth University
legitimacy of our field by pursuing an outmoded model—that “real” diseases are monocausal. School of Medicine, Richmond.
Despite ample evidence to the contrary, monocausal thinking continues to influence our field, Corresponding Author: Kenneth S.
for example, in the popular but improbable view that we can, with a few key advances, move Kendler, MD, Virginia Institute for
easily from descriptive to etiologically based diagnoses. Psychiatric and Behavioral Genetics,
Department of Psychiatry, Virginia
Commonwealth University,
JAMA Psychiatry. doi:10.1001/jamapsychiatry.2019.1200 PO Box 980126, Richmond, VA
Published online June 19, 2019. 23298-0126 (kenneth.kendler@
vcuhealth.org).

(aka “psychological”). The leading causes of each were, respec-

Many Causes
Robert Burton, an Oxford don who himself experienced depres-
sion, completed the first edition of his magnum opus The Anatomy
of Melancholy in 1621. In this encyclopedic book, he summarized the
extant theories of the “kinds, causes, symptoms, prognostics and
several cures” for melancholy.1 It is his list of the causes that are of
interest to the present article (Table 1). They number 44 and are di-
verse, including old age, temperament, parents, poverty, over-
much alcohol, sorrow, anger, immoderate eating, and envy.
In 1786, at the beginning of “modern” psychiatry, the British
alienist Thomas Arnold provided a detailed list of bodily and men-
tal causes of “insanity” (Table 2 and Table 3).2 Although lacking the
supernatural elements of Burton’s list of causes, Arnold’s list is even
more extensive, including dryness of the brain, inflammation, con-
cussion, ambition, jealousy, and despair.
In the opening chapter of his 1838 text, Jean Esquirol, the pre-
eminent French alienist of his age, provided tables of the most com-
mon causes of insanity (“La Folie”) recorded at a major Parisian pub-
lic asylum (Le Salpêtrière) and at his private madhouse (Table 4).3,4
Like Arnold, he divides the causes into 2 groups: physical and “moral”
tively, heredity and convulsions during pregnancy, and domestic
troubles and disappointed affection.
Such diverse lists of causes were not unique to mental illness.
In reviewing descriptions of scurvy in the late 18th century, 5
Harrison quotes the typical views of Baron van Swieten, an Aus-
trian court physician, who regarded “scurvy as a disease brought
about by humours that had become putrid, acrimonious and con-
densed…. This meant avoiding external causes such as “noisome
vapours, arising from marshy grounds and stagnating waters, inac-
tion, scarcity of greens and vegetables, drinking of corrupted and
stagnating waters, the use of salted and smoked flesh and
fish…damp and low lodgings.” It also meant counteracting predis-
posing factors such as “fear and sorrow” and avoiding food which
had gone bad.”5(p9)
Smith6 summarized the etiologic approach to cholera, specifi-
cally, and fevers, more generally, in the 1830s: “The causes…were
often discussed in terms of predisposing and exciting causes. In gen-
eral, predisposing causes were factors such as inadequate diet, over-
work, poverty, inadequate housing or ventilation, debilitation
through alcohol and mental exhaustion. The exciting causes were
those which drove the debilitation due to predisposing causes

jamapsychiatry.com (Reprinted) JAMA Psychiatry Published online June 19, 2019 E1

© 2019 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Idaho State University User on 07/18/2019

 Clinical Review & Education Special Communication From Many to One to Many—the Search for Causes of Psychiatric Illness

Table 1. Causes of Melancholia in Burton’s Table 3. Moral (aka Psychological) Causes of Insanity
The Anatomy of Melancholy, 1621a as Described by Arnold, 1786a

General Cause Individual Causes Category of Cause Individual Causes

Supernatural God, devil, witches Intense application Study; business or schemes or much exertion of genius;
of the mind employment of the mind, keeping it for a long time
Natural in an active wakeful state
General Passions of various Surprise or astonishment; joy; enthusiasm or religious
Inward Stars, old age, temperament, parents (hereditary) kinds when violent joy; pride and vanity; desire, anger, hatred or aversion,
or habitual love, ambition; avarice, distress, grief, despair, fear,
Outward Nurses, education, terrors, scoffs, imprisonment, suspicion, jealousy, anxiety, religious fear
poverty, losses Too great activity
Specific of imagination
Head Imbecility
Inward Humors, hot brain, excess venery, agues, stomach a
See Arnold.2
fumes
Outward Heat, blow to the head, overmuch alcohol, idleness
Windy Table 4. Physical and Moral (aka Psychological) Causes
Inward Default of bowel, ceased bleeding hemorrhoids of Insanity as Described by Esquirol, 1838a
Over body No. of Cases
Inward Distempered liver, bad diet Public Asylum Le Esquirol’s
Salpêtrière, Private
Diet Coarse bread, unclean water or milk, sharp sauces, Causes 1811-1812 Establishment
immoderate eating
Physical
Necessary Constipation; dark, thick air; excessive exercise;
over much or over little sleep Heredity 105 150
Passions Irascible, sorrow; fear; shame, disgrace; envy, malice; Convulsions during pregnancy 11 4
hatred, desire for revenge; anger; miseries;
concupiscible; ambition; covetousness; excess Epilepsy 11 2
of pleasures; pride, desire of praise; excess study Menstrual disorder 55 19
See Burton.1 Results of confinement 52 21
Critical period 27 11
Progress of age 60 4
Table 2. Bodily Causes of Insanity as Described by Arnold, 1786
Isolation 12 4
Category of
Cause Individual Causes Blows or falls upon the head 14 4
Internal Firmness, hardness, dryness of the brain; hard substances, Fevers 13 12
causes in tumors, excrescences; purulent mucous matter, blood, water
the brain in the ventricles; distension or enlargement; inflammation; Syphilis 8 1
indurations, adhesions; other causes of compression, irritation Mercury 14 18
External Exostosis of the skull; depression, fracture of the skull; Intestinal worms 24 4
causes on concussion; burning rays of the sun on the head; smallness
the brain of the head compressing the brain; swelling in the neck Apoplexy 60 10
Causes Phrenitis, delirious fevers, small pox; immoderate exercise; Moral
affecting profuse evacuations; excessive venery; defect of nourishment;
the body in other fevers; inactive, sedentary life Domestic troubles 105 31
general by Disappointed affection 46 25
acting
directly on Political events 14 31
the brain
Fanaticism 8 1
Causes Disease of abdominal viscera;a causing infirm state of the
acting in stomach or intestines;b active substances taken into the Fright 38 8
particular stomach: wine, opium, narcotics, or poisons; disordered Jealousy 18 14
body parts, state of womb, ovaries, or spermetic vessels; retention of
acting customary evacuation: lochia, milk, menses, piles, old ulcers; Anger 16 0
speedily or metastases to brain of gout, erysipelas, herpes, or cutaneous Misery and reverses of fortune 77 14
slowly on eruption; worms in the nostrils or frontal sinuses
the brain Wounded self-love 1 16
a 2
Arnold lists the following specific body parts: stomach, pylorus, intestines, Disappointed ambition 0 12
pancreas, mesentery, mesenteric glands, liver, spleen, hemorrhoidal vessels, Excessive study 0 13
kidneys, glandulae atrabiliariae, omentum, and peritoneum.
Misanthropy 0 2
b
Hard and indigestible aliment, immoderate use of warm diluting liquors,
a
intemperance of diet, viscous phlegm, and worms. See Esquirol.3,4

towards a particular pathological form, such as the specific atmo-
spheric factors seen as leading to cholera outbreaks.”6(p921)
As noted by Carter,7 prior to the last half of the 19th century,
causes of illness were anecdotally recorded by physicians from in-
dividual cases, resulting in an accumulation of long and diverse lists
for each disease. He proceeds: “For example, in his account of dia-
betes, Bardsley (1845) identified the following causes: frequent ex-
posure to sudden alterations of heat and cold, indulgence in copi-
ous drafts of cold fluid when the system has been over-heated by
labor or exercise, intemperate use of spiritous liquors, poor living,
sleeping of the whole of the night in the open in a state of intoxica-
tion, checking perspirations suddenly and mental anxiety and dis-
tress. Similar lists can be found for virtually any disease in most
German, English, or French medical texts from the period.”7(p11)

E2 JAMA Psychiatry Published online June 19, 2019 (Reprinted) jamapsychiatry.com

© 2019 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Idaho State University User on 07/18/2019

 From Many to One to Many—the Search for Causes of Psychiatric Illness
Toward One Cause
All of this changed in the latter half of the 19th century. Due largely
to the work of 2 men, Pasteur and Koch, the causal paradigm of medi-
cine shifted dramatically to what Dubos8 termed the doctrine of
specific etiology. He writes, “There is no more spectacular phenom-
enon in the history of medicine than the rapidity with which the germ
theory of disease became accepted by the medical profession-
.…From the haphazard accumulation of clinical facts, these clini-
cians derived generalizations that made it possible to classify dis-
eases as if they were defined biological entities.…[C]linical analysis
and classification [were developed and]…managed to make order
of the incredibly confused field of febrile diseases…. [D]iphtheria and
scarlatina emerged as well-defined diseases out of the chaos of the
‘fevers.’”8(p325)
The distinguished epidemiologist Mervyn Susser9 suggested
that the startling developments in infectious disease in the late 19th
century “led to the redefinition and reclassification of many dis-
ease entities by the criterion of cause…. By current definition, tu-
berculosis is caused by the tubercle bacillus.”9(p23) That is, the mono-
causal doctrine of specific etiology fundamentally changed the way
medical diagnoses were conceptualized. The desired approach was
no longer descriptive. Rather, the task was, whenever possible,
to define a disorder in relation to its 1 specific underlying cause.
Disorders not amenable to such causal clarification were of lower
status—syndromes and not real diseases. The monocausal theory—
derived from work on bacteria and viruses—soon spread to include
parasitic and vitamin-deficiency diseases.7
Psychiatry sought to participate in this diagnostic revolution
characterized by Dubois as the rise of the “doctrine of specific
etiology.”8,10 In the same year as his influential monograph on he-
bephrenia (1871),11 Hecker published a detailed defense of the di-
agnostic approach to psychiatric disorders of his mentor Kahlbaum:
“Just as in somatic pathology where prognosis and treatment be-
came infinitely more secure after the clinical disease forms fever,
typhoid fever, etc, branched off from the general symptom constel-
lation fever, the definition of clearly demarcated psychological dis-
ease forms, established according to the principles outlined above,
will greatly inform our diagnosis, prognosis and therapy, and help
us to shed the current uncertainty.”12(p353)
As general medicine turned from generic labels such as “fe-
vers” to disorders with specific etiologies, courses, preventions and
cures (the latter 2 by vaccination and serum treatments, respec-
tively), so too did psychiatry, under the influence of Kahlbaum and
Hecker’s teachings, later largely adopted by Kraepelin.13 This para-
digm sought to turn symptom-based diagnoses into distinct dis-
eases discovered using an iterative clinical research paradigm of signs,
symptoms, and course of illness. Implicit in this model (as it was in
the advances in microbiology that it imitated) was that the resul-
tant diseases so defined would be monocausal.
In this quest, the young field of psychiatry was greatly aided by a
twist of fate. In the mid-to-late 19th century, the central nervous sys-
tem infection by Treponema pallidum produced a syndrome, general
paresis of the insane (GPI), which as documented by Kraepelin ac-
counted for around 5% of asylum admissions in rural districts in Ger-
many and to up to 45% in urban public hospitals.14 Although the de-
finitive infectious etiology of GPI was not proven until early in the 20th
jamapsychiatry.com
© 2019 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Idaho State University User on 07/18/2019
Special Communication Clinical Review & Education
century,15 the distinctiveness of the syndrome—its specific neurologic
signs and rapid deteriorative course—were well recognized by the mid-
19th century. For example, in describing his research program for turn-
ing psychiatric syndromes in diseases, Hecker saw GPI as the paradig-
maticsuccessfulexample:“Theonlyrealdiseaseform,whichhasgained
entry into psychiatry, provides the best evidence for this: the so-called
progressive paralysis of the insane [AKA GPI].”12(p353) Many other de-
scriptions in the late 19th century by prominent neuropsychiatrists, in-
cludingKraepelin,documenttheparadigmaticroleplayedbyGPIinthe
search for monocausal psychiatric disorders (eg, see Kahlbaum,16(p27)
Behr,17(p5) Daraszkiewicz,18(p10) and Kraepelin19(p132)).
Histories of Epidemiology and Biostatistics—
From One to Many
The histories of 2 additional fields need to be interwoven into our
narrative: epidemiology and statistics. Let us first examine epide-
miology. Susser and Susser, using the Kuhnian concept of a dominant
scientific paradigm, described 3 successive eras of epidemiology.20
The first focused on sanitary statistics, had as its paradigm “mi-
asma,” and dominated the field until the mid-19th century. Miasma
frequently occurred in the long lists of causes of illnesses in pre-
1850s medicine but never as the single cause. The second epide-
miologic phase, termed infectious disease, had as its paradigm “the
germ theory,” lasted roughly from 1850 to 1950, and was domi-
nated by monocausal etiologic theories in which the relationship be-
tween putative etiologic agents and specific diseases was one to one
(although most investigators recognized such modifying factors as
“host resistance”). The third epidemiological phase, termed chronic
disease, had a dominant “black box” paradigm. It incorporated a mul-
tifactorial disease model for what was termed chronic noncommu-
nicable diseases (eg, diabetes, heart disease, certain forms of can-
cer, and hypertension), diseases often associated with particular
lifestyles that could not be explained by a single salient causal fac-
tor. In this paradigm, the goal of epidemiology was to determine the
magnitude and causal nature of the associations between a wide
range of putative risk factors and these chronic noncommunicable
diseases. This phase began around 1950 and has lasted until cur-
rent times.
Two metaphors well capture this third phase of medical epide-
miology. The first of these is the “causal pie” proposed by Rothman.21
He writes, “The causal pie model posits that several causal compo-
nents act in concert to produce an effect…. One can view each causal
pie as a set of interacting causal components.21(pS145)
The second metaphor, a “web of causation,” is described by
Krieger22: “Conceptually, the metaphor evoked the powerful im-
age of a spider’s web, an elegantly linked network of delicate strands,
the multiple intersections representing specific risk factors or out-
comes, and the strands symbolizing diverse causal pathways. It en-
couraged epidemiologists to look for multiple causes and multiple
effects, and to identify the many—as opposed to singular—routes by
which disease could be prevented.”22(p891)
Parallel to these changes in epidemiology, statistical and proba-
bilistic approaches toward diseases evolved. As argued by Gillies,23
based on a detailed study of Koch’s postulates (see also Carter7):
“19th and early 20th century scientific medicine, as it was devel-
oped by Pasteur, Koch, and others, used a deterministic notion of
(Reprinted) JAMA Psychiatry Published online June 19, 2019 E3
 Clinical Review & Education Special Communication
causality. An attempt was made to show that each disease had a
single cause which was both necessary and sufficient.”23(p2)
However, around 1950, indeterministic causal models became in-
creasinglyimportant.Manyfactorsinfluencedthischange,butonewas
predominant—theintensedebateaboutandeventualresolutionofthe
causal association between smoking and lung cancer. This debate led
to the proposal and eventual wide acceptance of the Hill criteria24 for
causation. Smoking became a paradigmatic risk factor that was an im-
portant cause of lung cancer despite unequivocal evidence that many
people who were long-term smokers never developed lung cancer and
many individuals with lung cancer never smoked.
At this time, biostatistics was developing the statistical machin-
ery to detect and further characterize a wide range of causes whose
relationship with their outcome was probabilistic. Rothman and
Greenland25 made a key point regarding the public health signifi-
cance of such causes: “The importance of multicausality is that most
identified causes are neither necessary nor sufficient to produce dis-
ease. Nevertheless, a cause need not be either necessary or suffi-
cient for its removal to result in disease prevention. If a component
cause that is neither necessary nor sufficient is blocked, a substan-
tial amount of disease may be prevented.”25(p145)
The Anomalous History of Late 20th Century
Psychiatry—Focused on Monocausal Theories
During the second half of the 20th century, the approach to dis-
ease causation of major parts of psychiatry was out of step with the
rest of medicine and medical epidemiology. Instead of multicausal
models, the rising and soon to be dominant field of biological psy-
chiatry pursued monocausal models for their major disorders.
This search has 2 prominent phases, both fueled by new scientific
developments. The first was neurochemical. The stage was set
in 1957 by Montagu’s discovery of dopamine in brain tissue26
quickly followed, in 1960, by the dramatic finds from Ehringer
and Hornykiewicz of the decreased content of dopamine in the post-
mortem brains of patients with Parkinson disease.20 Here was a
major neurologic disorder fitting apparently into a monocausal neu-
rochemical theory. What could be more exciting for the then young
and ambitious field of biological psychiatry? Even greater excitement
followed the impressive treatment effects obtained for L-dopa (leva-
dopa)inpatientswithParkinsondisease.27 Thus,thesinglemajorcause
predicted a viable treatment.
In the early to mid-1960s, through histofluorescence
stains, the cell bodies and neuronal pathways of the 3 monoamine
neurotransmitters were identified: dopamine, norepinephrine,
and serotonin.28-31 This further spurred the development of 3
long-lived monocausal neurochemical theories for psychiatric ill-
ness. All were proposed in the mid-to-late 1960s: the “catechol-
amine hypothesis of affective disorders,” 32,33 the “dopamine
hypothesis of schizophrenia,”34,35 and the “serotonin hypothesis
of depression.”36 Although based on a range of evidence, the pri-
mary support for these theories was reasoning backward from
therapeutic mechanisms to etiology. That is, for Parkinson disease
the logic was sound: clarification of cause leading to a proposed
treatment. By contrast, psychiatry followed the more problematic
approach of extrapolating backward from a proposed mechanism
of treatment to the cause of the disease. Although the original
E4 JAMA Psychiatry Published online June 19, 2019 (Reprinted)
© 2019 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Idaho State University User on 07/18/2019
From Many to One to Many—the Search for Causes of Psychiatric Illness
articles proposing these theories were couched in qualifications,
as a psychiatry resident in the late 1970s, I was taught these theo-
ries as monocausal explanations. Schizophrenia was caused by
excess dopamine transmission. Decades later, I would commonly
see patients who would say some version of “my psychiatrist said
I have a chemical imbalance in my brain” and then proceed to sum-
marize one or more of these theories. It is now widely accepted
that these theories, claiming a dominant causal pathway to illness,
are false although debate continues regarding the dopamine
hypothesis.37,38
The second wave of monocausal theories in psychiatry was ge-
netic. Despite much evidence from family studies that major psy-
chiatric disorders did not segregate in pedigrees as expected for a
mendelian condition, the first successful linkage study of Hunting-
ton disease in 198339 elicited intense excitement in psychiatric ge-
netics and launched a large number of linkage studies, especially of
schizophrenia and bipolar illness. These early investigations used
small numbers of pedigrees and were only powered to detect genes
of a mendelian-like effect. The common parlance in those days was
that the hunt was on for “the gene for…” This led to 3 reports in 1987
through 1988 in prominent journals claiming successful linkage to
genes of large effect for schizophrenia40 and bipolar illness.41,42 The
field thought that success was at hand. However, all 3 of these claims
were later disproved.
Thirty years later, we have published evidence of 108
schizophrenia-associated common loci in the human genome using
genome-wide association studies (GWAS),43 with unpublished re-
sults more than doubling that number. For major depression, the
current number is 4444 and increasing. Smaller number of loci are
identified for other disorders (eg, bipolar illness, autism, obsessive-
compulsive disorder, and anorexia), but the evidence suggests that
as sample sizes rise, the number of identified risk variants inexora-
bly increases.45 Molecular genetics provides us with the clearest ex-
ample of the “one to many” transition in causal theories of psychi-
atric illness in recent decades. We started by seeking the gene for
our major disorders by using linkage analysis and ended up finding,
for these same conditions, many genes of small effect with GWAS.
Current Evidence for the Multicausal Nature
of Psychiatric Illness
Anyone who reads leading psychiatry or clinical psychology jour-
nals on a regular basis will see that high-quality research on the causes
of psychiatric illness spans a wide array of perspectives and subdis-
ciplines. In 2013, I sought to rigorously assess this informal impres-
sion by reviewing all studies that addressed the etiology of psychi-
atric disorders in the first four 2013 issues of 12 major psychiatry and
psychology journals.46 In the resulting 197 reviewed articles, I rated
the putative causes that were examined using 12 categories. Five
categories reflected various levels of biological science: molecular
genetics; molecular, neurochemical, or cellular neuroscience; sys-
tems neuroscience; aggregate genetic effects; and miscellaneous
biological influences. Four categories reflected various levels of psy-
chological science: neuropsychological traits; personality and other
trait-like measures; cognitions or attitudes; and psychiatric symp-
toms or other disorders. Three categories reflected environmental
effects: individual; family; and community, society, or culture.
jamapsychiatry.com
 From Many to One to Many—the Search for Causes of Psychiatric Illness
I identified 306 individual predictor variables from those pa-
pers, and the variables were widely distributed across the 12 levels.
Our discipline, which includes individuals with expertise in molecu-
lar biology, neuroscience, genetics, imaging, cognition, personal-
ity, clinical and developmental psychology, epidemiology, and soci-
ology, has provided rigorous evidence of potential and widely diverse
causes of psychiatric illness. Although the diversity and number of
these causes resemble the lists generated by our predecessors in the
17th through 19th centuries, these causes are the product of scien-
tific study, not anecdotal bedside observations. However, of those
197 studies, only 22 (11%), engaged in meaningful, integrative, cross-
level research.
Seeking to close the circle that started with Burton’s list of causes
in his The Anatomy of Melancholy,1 I set myself the task of identify-
ing, either from review papers or replicated studies, risk factors that
are associated with major depression and that could plausibly be con-
sidered causes. I organized them in the categories used in my prior
article. As seen in the eTable in the Supplement, while far from en-
cyclopedic, I was able to identify 37 potential causes of depression
distributed across 11 levels of scientific inquiry, nearly equal to
Burton’s number of 44.
Implications
Psychiatry has had a long-term love affair with monocausal theories
of illness dating at least from the late 19th century, heavily influenced
by our success at the identification and effective eradication of GPI.
In the latter half of the 20th century, with both neurochemical and mo-
lecular genetic theories of illness, our enthusiasm for monocausal
theoriesoutranourcommonsense.Emergingfromdecadesofpsycho-
analytic dominance, we were deeply committed to reestablishing our
medical legitimacy. What better way to show that we treated “real” dis-
eases than to show that they were monocausal?
This is in retrospect ironic because such an approach fitted bet-
ter into the world of medicine in the late 19th century than the late
20th century. In those hundred years, Western medicine and medi-
cal epidemiology moved from a focus on monocausal infectious dis-
eases to the multifactorial chronic disease model. In the late 20th
century, it would have been more accurate for psychiatry to claim
that we treated “real” disorders because they were multicausal, simi-
lar to coronary artery disease, hypertension, rheumatoid arthritis,
and type 2 diabetes, not that they were monocausal, such as clas-
sic infectious or mendelian medical disorders.
Our long yearning for monocausal theories of etiology drives, at
least in part, our heartfelt calls for the abandonment of our descriptive
nosologic systems in favor of an etiologic model. Implicit in this posi-
tion is the expectation that we will have one definitive causal explana-
tion for our major disorders that operates at one scientific level, typi-
cally a biological one. That is, as a bacterial pneumonia can be shown
tobeattributabletoaspecificmicrobiologicalcauseoramendeliandis-
order a result from a definitive causal alteration in DNA sequence, so
we can give up our symptoms and signs and replace them by a defini-
tive etiologic laboratory test or evidence of an aberrant brain circuit.
This will not happen for major psychiatric disorders because they
have multiple levels of etiology; thus, there is not 1 possible etio-
logic model that could be used, but rather many.47 Although each
subdiscipline within psychiatry might favor its own particular ap-
jamapsychiatry.com
© 2019 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Idaho State University User on 07/18/2019
Special Communication Clinical Review & Education
proach (molecular genetics, systems neuroscience, neuropsychol-
ogy, environmental exposures, etc.), we have no good way to arbi-
trate between them. So rather than having a breakthrough moment
when we discover the definitive etiologic measure that can replace
our signs and symptoms, we should prepare for a scenario with
slower, more incremental progress.
Consider how laboratory tests are used in general medicine out-
side of microbiology or medical genetics. Typically, we find practi-
cal tests from different levels of explanations that help in our diag-
nostic task. We use the electrocardiogram and plasma troponin levels
to help us diagnose myocardial infarction not because the tests di-
rectly index the monocausal etiology of coronary atherosclerosis but
because they are practical and have good sensitivity and specificity
for detecting key consequences of the underlying disorder. But those
tests are not ordered unless the patient displays some features of
the classic symptoms and signs well known by any physician who
has spent time in an emergency department. Indeed, their sensi-
tivity and specificity make sense only when they are assessed in pa-
tients with a particular clinical presentation. Laboratory tests will
evolve in a similar way in psychiatry. We will not likely be giving
up our long-accumulated clinical diagnostic wisdom any time soon.
While disappointing to visions of dramatic progress in psychiatry,
giving up our dreams of monocausal theories will reflect increased
scientific maturity and realism.
Caveats
Acceptance of a multicausal model for psychiatric illness should be no
causeforcynicismaboutbiologicaletiologicresearch.Indeed,formany
highly multicausal medical disorders, effective treatments exist, some
ofwhichweredevelopedinresponsetoclarificationofparticularcausal
pathways. Equifinality is also a realistic possibility—where pathways to
illnessfromseveralseeminglydisparateriskfactorscoalescetoacausal
bottleneck at which interventions could be aimed.
Monocausal subforms of major psychiatric disorders remain to be
identified. We have discovered rare copy number variants that carry
a substantial risk for schizophrenia, but none have close to complete
penetrance.48 Some de novo genetic variants strongly predispose to
autism,especiallytheformwithlowIQandothersyndromalfeatures.49
But we have ruled out single major locus genetic causes where efforts
havebeenmadeinmanypartsoftheworldtolocatepedigreesinwhich
psychiatric disorders, especially schizophrenia and bipolar affective ill-
ness, segregate like a mendelian disorder. Although certainty here is
impossible, we have extensive evidence that suggests that our major
psychiatric disorders are multicausal “all the way down.”
Final Thoughts
Our journals are now filled with evidence for the multicausal nature
of psychiatric conditions.50 Thus, this matter should be settled. Yet
the ghost of GPI—of monocausal psychiatric disorders—lurks in our
memory. To this day, it influences our nosologic thinking. It makes
us too willing to adopt a monocausal perspective in our clinical work
and in our explanations of psychiatric disorders to patients. Mono-
causal thinking continues to support hard reductionist approaches
that seek the cause of our major disorders and is one of several fac-
(Reprinted) JAMA Psychiatry Published online June 19, 2019 E5
 Clinical Review & Education Special Communication From Many to One to Many—the Search for Causes of Psychiatric Illness

tors inhibiting collaborative psychiatric research work across
scientific levels.
Despite the wide acceptance of the chronic noncommuni-
cable disease model in modern medicine, there remains in our cul-
ture a sense that disorders with many causes have reduced legiti-
macy. Therefore, both clinicians and their patients would feel more
secure if a large indisputable cause were found for their disorders.
This, however, is a social and not a scientific problem. The stigma of
psychiatric illness and the low status of the psychiatric profession
need to be addressed at both social and political levels and will not
likely be solved through the discovery of major single causes for our
illnesses. The legitimacy of the discipline of psychiatry does not rest
on our ability to find single major causes of our disorders.
Some may mourn the passing of monocausal theories for psychi-
atric illness. I prefer to view the process more positively. Rather than
grieving for the loss of our visions of another GPI around the corner, we
can positively embrace the etiologic complexity of our disorders. After
all, as psychiatrists, we have chosen to study the most complex func-
tionsofthehumanbeing’smostintricateorgan.Thehumanbrain/mind
not only self-wires—a system far more intricate than the genome can
specify in any one-to-one relationship—but also is the great organizer
and compiler of our own existence. It remembers and changes in ways
that the liver, kidney, or heart can never do. It interfaces between our
organism and the psychological, social, and cultural world around us
to provide the structure and meaning of our existence. We have cho-
sentostudyandtreatthemostcomplexofhumandisorders,thecauses
of which span the many levels of our biology, our psychology, and our
socialexistence.Ifthecommon,morbiddysfunctionsofthehumancar-
diovascular, immune, hormonal, musculoskeletal, and gastrointesti-
nalsystems,whichcausemostofthemorbidityinourcountry,arehighly
multifactorial, could we realistically expect anything else from the par-
allel dysfunctions of our mind/brain system?

ARTICLE INFORMATION 9. Susser M. Causal Thinking in the Health Sciences. 23. Gillies D. Causality, Probability, and Medicine.
Accepted for Publication: April 7, 2019. New York, NY: Oxford University Press; 1973. London: Routledge, Taylor & Francis Group; 2019.

Published Online: June 19, 2019.
doi:10.1001/jamapsychiatry.2019.1200
Conflict of Interest Disclosures: None reported.
Additional Contributions: Peter Zachar, PhD,
Department of Psychology, Auburn University at
Montgomery, Alabama, and Joel Braslow, MD,
Department of Psychiatry, University of California,
Los Angeles, provided helpful comments on earlier
versions of the manuscript. They were not
financially compensated for their contributions.
The manuscript was stimulated in part by reading
the insightful essay of Ross.10
REFERENCES
1. Burton R. The Anatomy of Melancholy. What it is,
With All the Kinds, Causes, Symptomes,
Prognostickes, and Seuerall Cures of it. Oxford: John
Lichfield and James Short for Henry Cripps; 1621.
2. Arnold T. Observations on the Nature, Kinds,
Causes, and Prevention of Insanity, Lunacy, or
Madness. Leicester: Printed by G. Ireland, for
G. Robinson, 1786.
3. Esquirol E. Des Maladies Mentales. Considérées
Sous Les Rapports: Médical, Hygiénique et
Médico-Légal [in French]. Tome Premier. Bruxelles:
Librairie Médicale et Scientifique de J.B. Tircher,
1838.
4. Esquirol JED. Mental Maladies: Treatise on
Insanity. Hunt EK, trans. Philadelphia, PA: Lea and
Blanchard; 1845.
5. Harrison M. Scurvy on sea and land: political
economy and natural history, c. 1780-c. 1850.
J Marit Res. 2013;15(1):7-25. doi:10.1080/21533369.
2013.783167
6. Smith GD. Commentary: behind the Broad
Street pump: aetiology, epidemiology and
prevention of cholera in mid-19th century Britain.
Int J Epidemiol. 2002;31(5):920-932. doi:10.1093/
ije/31.5.920
7. Carter KC. The Rise of Causal Concepts of
Disease: Case Histories. Burlington, VT: Ashgate
Publishing Co; 2003.
8. Dubos RJ. Man Adapting. New Haven, CT:
Yale University Press; 1980.
10. Ross LN. The doctrine of specific etiology. Biol
Philos. 2018;33(37):1-22.
11. Hecker E. Die hebephrenie: ein beitrag zur
klinischen psychiatrie. Virchows Archiv für
pathologische Anatomie und Physiologie und für
Klinische Medizin. 1871;52:394-429. doi:10.1007/
BF02329963
12. Kraam A. On the origin of the clinical standpoint
in psychiatry, Dr Ewald Hecker in Görlitz. Hist
Psychiatry. 2004;15(59, pt 3):345-360. doi:10.1177/
0957154X04044598
13. Kendler KS, Engstrom EJ. Kahlbaum, Hecker,
and Kraepelin and the transition from psychiatric
symptom complexes to empirical disease forms.
Am J Psychiatry. 2017;174(2):102-109. doi:10.1176/
appi.ajp.2016.16030375
14. Kraepelin E. General Paresis. Moore JW, trans.
New York, NY: The Journal of Nervous and Mental
Disease Publishing Co; 1913.
15. Noguchi H, Moore JW. A demonstration of
treponema pallidum in the brain in cases of general
paralysis. J Exp Med. 1913;17(2):232-238. doi:10.
1084/jem.17.2.232
16. Kahlbaum KL. Catatonia. Levij Y and Pridan T,
trans. Baltimore, MD: Johns Hopkins University Press;
1973.
17. Behr A. Die Frage der “Katatonie”, oder
des Irreseins mit Spannung [in German].
Univerisitas Dorpatensis: Riga: WF Häcker; 1891.
18. Daraszkiewicz L. Ueber hebephrenie,
insbesondere deren schwere form:
inaugural-dissertation. Dorpat: Laakmann's Buch
und Steindruckerei; 1892.
19. Kraepelin E. Dementia Praecox and Paraphrenia.
Huntington, NY: Krieger Publishing; 1971.
20. Hornykiewicz O. The discovery of dopamine
deficiency in the parkinsonian brain. J Neural
Transm Suppl. 2006;(70):9-15. doi:10.1007/978-3-
211-45295-0_3
21. Rothman KJ. Causes. Am J Epidemiol. 1976;104
(6):587-592. doi:10.1093/oxfordjournals.aje.a112335
22. Krieger N. Epidemiology and the web of
causation: has anyone seen the spider? Soc Sci Med.
1994;39(7):887-903. doi:10.1016/0277-9536(94)
90202-X
24. Hill AB. The environment and disease:
association or causation? Proc R Soc Med. 1965;58
(4):295-300.
25. Rothman KJ, Greenland S. Causation and causal
inference in epidemiology. Am J Public Health.
2005;95(suppl 1):S144-S150. doi:10.2105/AJPH.
2004.059204
26. Montagu KA. Catechol compounds in rat
tissues and in brains of different animals. Nature.
1957;180(4579):244-245. doi:10.1038/180244a0
27. Kopin IJ. Parkinson’s disease: past, present,
and future. Neuropsychopharmacology. 1993;9(1):
1-12. doi:10.1038/npp.1993.39
28. Falck B, Hillarp NA, Thieme G, Torp A.
Fluorescence of catechol amines and related
compounds condensed with formaldehyde. Brain
Res Bull. 1982;9(1-6):xi-xv. doi:10.1016/0361-9230
(82)90113-7
29. Dahlstroem A, Fuxe K. Evidence for the
existence of monoamine-containing neurons in the
central nervous system, I: demonstration of
monoamines in the cell bodies of brain stem
neurons. Acta Physiol Scand Suppl. 1964;232:232,
1-55.
30. Fuxe K, Hökfelt T, Nilsson O, Reinius S.
A fluorescence and electron microscopic study on
central monoamine nerve cells. Anat Rec. 1966;155
(1):33-40. doi:10.1002/ar.1091550105
31. Dahlstroem A, Fuxe K, Olson L, Ungerstedt U.
Ascending systems of catecholamine neurons from
the lower brainstem. Acta Physiol Scand. 1964;62:
485-486. doi:10.1111/j.1748-1716.1964.tb10446.x
32. Schildkraut JJ. The catecholamine hypothesis
of affective disorders: a review of supporting
evidence. 1965. J Neuropsychiatry Clin Neurosci.
1995;7(4):524-533. doi:10.1176/jnp.7.4.524
33. Hirschfeld RMA. History and evolution of the
monoamine hypothesis of depression. J Clin
Psychiatry. 2000;61(suppl 6):4-6.
34. van Rossum JM. The significance of
dopamine-receptor blockade for the mechanism of
action of neuroleptic drugs. Arch Int Pharmacodyn
Ther. 1966;160(2):492-494.
35. Kendler KS, Schaffner KF. The dopamine
hypothesis of schizophrenia: an historical and

E6 JAMA Psychiatry Published online June 19, 2019 (Reprinted) jamapsychiatry.com

© 2019 American Medical Association. All rights reserved.

Downloaded From: https://jamanetwork.com/ by a Idaho State University User on 07/18/2019

 From Many to One to Many—the Search for Causes of Psychiatric Illness
philosophical analysis. Philos Psychiatry Psychol.
2011;18(1):41-63. doi:10.1353/ppp.2011.0005
36. Coppen A. The biochemistry of affective
disorders. Br J Psychiatry. 1967;113(504):1237-1264.
doi:10.1192/bjp.113.504.1237
37. Edwards AC, Bacanu SA, Bigdeli TB, Moscati A,
Kendler KS. Evaluating the dopamine hypothesis of
schizophrenia in a large-scale genome-wide
association study. Schizophr Res. 2016;176(2-3):
136-140. doi:10.1016/j.schres.2016.06.016
38. Gründer G, Cumming P. The dopamine
hypothesis of schizophrenia: current status. In: Abel
T, Nickl-Jockschat T, eds. The Neurobiology of
Schizophrenia. Cambridge, MA: Academic Press;
2016:109-124. doi:10.1016/B978-0-12-801829-3.
00015-X
39. Gusella JF, Wexler NS, Conneally PM, et al.
A polymorphic DNA marker genetically linked to
Huntington’s disease. Nature. 1983;306(5940):
234-238. doi:10.1038/306234a0
40. Sherrington R, Brynjolfsson J, Petursson H,
et al. Localization of a susceptibility locus for
schizophrenia on chromosome 5. Nature. 1988;336
(6195):164-167. doi:10.1038/336164a0
41. Egeland JA, Gerhard DS, Pauls DL, et al. Bipolar
affective disorders linked to DNA markers on
jamapsychiatry.com
© 2019 American Medical Association. All rights reserved.
Downloaded From: https://jamanetwork.com/ by a Idaho State University User on 07/18/2019
Special Communication Clinical Review & Education
chromosome 11. Nature. 1987;325(6107):783-787.
doi:10.1038/325783a0
42. Baron M, Risch N, Hamburger R, et al. Genetic
linkage between X-chromosome markers and
bipolar affective illness. Nature. 1987;326(6110):
289-292. doi:10.1038/326289a0
43. Schizophrenia Working Group of the
Psychiatric Genomics Consortium. Biological
insights from 108 schizophrenia-associated genetic
loci. Nature. 2014;511(7510):421-427. doi:10.1038/
nature13595
44. Wray NR, Ripke S, Mattheisen M, et al;
eQTLGen; 23andMe; Major Depressive Disorder
Working Group of the Psychiatric Genomics
Consortium. Genome-wide association analyses
identify 44 risk variants and refine the genetic
architecture of major depression. Nat Genet. 2018;
50(5):668-681. doi:10.1038/s41588-018-0090-3
45. Sullivan PF, Agrawal A, Bulik CM, et al;
Psychiatric Genomics Consortium. Psychiatric
genomics: an update and an agenda. Am J Psychiatry.
2018;175(1):15-27. doi:10.1176/appi.ajp.2017.17030283
46. Kendler KS. The structure of psychiatric
science. Am J Psychiatry. 2014;171(9):931-938.
doi:10.1176/appi.ajp.2014.13111539
(Reprinted) JAMA Psychiatry Published online June 19, 2019
47. Kendler KS. Levels of explanation in psychiatric
and substance use disorders: implications for the
development of an etiologically based nosology.
Mol Psychiatry. 2012;17(1):11-21. doi:10.1038/mp.
2011.70
48. Marshall CR, Howrigan DP, Merico D, et al;
Psychosis Endophenotypes International
Consortium; CNV and Schizophrenia Working
Groups of the Psychiatric Genomics Consortium.
Contribution of copy number variants to
schizophrenia from a genome-wide study of 41,321
subjects. Nat Genet. 2017;49(1):27-35. doi:10.1038/
ng.3725
49. Ramaswami G, Geschwind DH. Genetics of
autism spectrum disorder. Handb Clin Neurol. 2018;
147:321-329. doi:10.1016/B978-0-444-63233-3.
00021-X
50. Paulus MP, Thompson WK. The challenges and
opportunities of small effects: the new normal in
academic psychiatry [published online February 27,
2019]. JAMA Psychiatry. doi:10.1001/
jamapsychiatry.2018.4540
E7