Biological Psychiatry

Abstract

Biological psychiatry is a branch of psychiatry that treats mental disorders as disturbances of brain function and seeks their causes in the biochemistry, circuitry, and genetics of the nervous system. Emerging from nineteenth-century nosology and transformed by the mid-century discovery that drugs altering neurotransmitters could relieve psychiatric symptoms, it built successive biological theories of mental illness, from the monoamine hypotheses of depression and psychosis to the polygenic genomics of the present. This article traces that arc: the neurochemical hypotheses and the receptor pharmacology behind them, the shift from molecules to neural circuits, the genome-wide reconstruction of psychiatric risk, and the enduring debate over how brain-based a psychiatric diagnosis can be. Three interactive demonstrations let the reader model receptor occupancy, polygenic liability, and dimensional classification.

Keywords: dopamine hypothesis, psychiatric genomics, biological psychiatry, research domain criteria

Biological psychiatry is the approach to mental illness that locates its causes in the physical operation of the brain and body and pursues its treatment by physical means. It is less a single theory than a research program with a working premise: that disturbances of mood, thought, and behavior are, at some level of description, disturbances of neural function, and that the tools of biochemistry, pharmacology, neuroimaging, and genetics can therefore illuminate them. That premise has a long pedigree. The German psychiatrist Emil Kraepelin, sorting the disorders of the asylum at the turn of the twentieth century, insisted that mental illnesses were natural disease entities with a biological course and outcome, a nosological instinct that distinguished dementia praecox from manic-depressive insanity and set the template modern classification still follows (Kraepelin, 1899). What turned that instinct into an empirical science was a pharmacological accident: the mid-twentieth-century discovery that specific drugs could specifically alter psychiatric symptoms, which implied that the symptoms had a chemistry to alter.

What Biological Psychiatry Is

Biological psychiatry is best understood as a stance toward explanation rather than a fixed body of doctrine. Its defining move is to ask, of any psychiatric phenomenon, what is happening in the brain that produces it, and to treat that question as answerable by the ordinary methods of the life sciences. This does not commit the field to the claim that biology is the whole story of mental illness; most of its practitioners hold a version of the biopsychosocial view, in which biological, psychological, and social causes interact. What the biological program contributes is the insistence that the biological level is real, measurable, and clinically consequential, and that ignoring it leaves psychiatry without a mechanism. The methodological core is a form of reductionism, the strategy of explaining a complex phenomenon by the properties of its parts, pursued here from the syndrome down through the circuit and the synapse to the gene. The neuroscientist Eric Kandel gave the program its most influential modern charter, arguing that all mental processes are brain processes, that even the effects of psychotherapy and social experience are ultimately changes in the brain, and that psychiatry's future lay in integrating this biology with the psychology it must still explain (Kandel, 1998). Table 1 sets out the successive frameworks through which the field has pursued that program, each layered onto the last rather than simply replacing it; the chapters that follow trace them in turn, beginning with the chemistry that started it.

The Monoamine Hypotheses

The first biological theories of mental illness were read backward off the actions of drugs. In the 1950s, clinicians noticed that chlorpromazine calmed psychosis, that reserpine, used for hypertension, could precipitate depression, and that iproniazid, a tuberculosis drug, lifted mood. Each of these agents was found to act on the monoamine neurotransmitters, the family that includes dopamine, norepinephrine, and serotonin, small signaling molecules released across the synapse to excite or inhibit the next neuron. The pharmacologist Arvid Carlsson supplied the decisive mechanism: administering the antipsychotics chlorpromazine and haloperidol to mice raised the brain levels of dopamine's metabolites, exactly what would happen if the drugs blocked dopamine receptors and the neuron compensated by releasing more transmitter (Carlsson & Lindqvist, 1963). If blocking dopamine relieved psychosis, perhaps psychosis was a state of excess dopamine signaling, and the dopamine hypothesis of schizophrenia was born. The same logic ran in the other direction for mood. Joseph Schildkraut synthesized the scattered pharmacological clues into the catecholamine hypothesis of affective disorders, proposing that depression is associated with a functional deficiency of norepinephrine at key synapses and mania with its excess, a theory that dominated biological thinking about mood for a generation (Schildkraut, 1965). These monoamine hypotheses launched modern psychopharmacology, the study of how drugs act on the mind through the brain, and gave the field its first testable bridge from a molecule to a syndrome. The demonstration below models the pharmacological heart of the dopamine hypothesis: how the fraction of receptors a drug occupies rises with its dose.

From Molecules to Circuits

The monoamine hypotheses were powerful but too simple, and their limits pushed the field upward from the synapse to the circuit. Depleting monoamines does not reliably cause depression in healthy people, and antipsychotics occupy their receptors within hours while taking weeks to work, so the chemistry could not be the whole mechanism. The decisive tool for the next level was neuroimaging, the family of techniques that visualize brain structure and activity in living patients. Nancy Andreasen, an early champion of applying imaging to psychiatry, argued that the discipline needed a scientific psychopathology that linked the symptoms of mental illness to measurable abnormalities of brain structure and function, using imaging to test whether disorders such as schizophrenia involved identifiable circuits rather than a diffuse chemical imbalance (Andreasen, 1997). Imaging reframed disorders as dysfunctions of distributed neural systems: the mesolimbic dopamine pathway in psychosis, prefrontal-limbic circuits in mood and anxiety disorders, connections that reach the very regions cognitive psychology studies for reward processing and executive control in the prefrontal cortex. It also created a way to connect genes to behavior across the intervening levels. Daniel Weinberger and Andreas Meyer-Lindenberg proposed studying intermediate phenotypes, measurable brain traits that lie between a risk gene and a clinical diagnosis and are simpler and more directly biological than the diagnosis itself; such an endophenotype, a heritable brain-level marker of vulnerability, could reveal a genetic effect on the brain long before it produced overt illness (Meyer-Lindenberg & Weinberger, 2006). Figure 1 lays out these levels of explanation as the field now conceives them, from the genome to the presenting symptom.

The Genomics of Psychiatric Disorders

If mental disorders are brain disorders, and if they run in families, then their risk should be written partly in the genome, and finding it became biological psychiatry's largest modern enterprise. Twin studies had long established that the major psychiatric disorders are substantially heritable, meaning that a large share of the variation in who develops them tracks genetic differences between people, with schizophrenia and bipolar disorder among the most heritable. The first attempt to locate the responsible genes, the candidate-gene era, chose plausible genes in neurotransmitter pathways and tested them one at a time; almost none of its findings replicated. What replaced it was the genome-wide association study, a method that scans millions of DNA variants across the whole genome in very large samples without any prior guess about which genes matter. Patrick Sullivan and colleagues, reviewing the early results, showed that the genetic architecture of psychiatric disorders is highly polygenic: risk is not carried by a few genes of large effect but distributed across thousands of common variants each raising risk almost imperceptibly (Sullivan et al., 2012). The many tiny effects can be summed for an individual into a polygenic score, a single number estimating genetic liability from the whole genome. A striking further result was that this risk is partly shared across diagnoses: the Cross-Disorder Group of the Psychiatric Genomics Consortium found specific genetic loci with effects spanning schizophrenia, bipolar disorder, major depression, autism, and attention-deficit disorder, undercutting the assumption that these are genetically distinct entities (Cross-Disorder Group of the Psychiatric Genomics Consortium, 2013). As samples grew, the schizophrenia analyses in particular matured, implicating hundreds of loci and, at last, coherent biology: the largest such study mapped 287 risk regions and pointed squarely at genes concentrated in the synapse (Trubetskoy et al., 2022). The demonstration below shows how thousands of small genetic effects combine into a liability that crosses a threshold into disorder.

Classification and Its Discontents

Biological psychiatry's deepest problem is that its biology has never mapped cleanly onto its diagnoses. The categories clinicians use, codified in the Diagnostic and Statistical Manual, are defined by lists of symptoms, not by any biological test, and a century of research has failed to find a marker that identifies a DSM disorder the way a blood test identifies diabetes. Steven Hyman, a former director of the National Institute of Mental Health, argued that the DSM categories had hardened into a reification, a convenient descriptive fiction mistaken for a natural kind, and that treating these committee-drawn boxes as though they were real diseases had actively impeded the search for their biology (Hyman, 2010). The genomic finding that risk is shared across supposedly distinct disorders sharpened the complaint. The institute's response, led by Thomas Insel, was the Research Domain Criteria, a framework that sets the DSM categories aside for research purposes and instead studies dimensions of function, such as reward, fear, and cognitive control, each analyzed across levels from genes to circuits to behavior, on the bet that biology will respect these dimensions even where it ignores the diagnoses (Insel et al., 2010). The dimensional turn reconnects psychiatry to the constructs of cognitive science and to processes studied in ordinary decision making, since a dimension like reward sensitivity is a psychological variable before it is a psychiatric one. The clinical synthesis for the field's flagship disorder captures the current consensus: schizophrenia is best understood as a neurodevelopmental disorder of polygenic risk expressed through disrupted brain circuits, no single level sufficient on its own (Owen et al., 2016). The demonstration below contrasts the categorical and dimensional views of the same population.

Worked Example

The pharmacological core of the dopamine hypothesis can be made quantitative with a standard model of drug binding. When a drug and its receptor reach equilibrium, the fraction of receptors the drug occupies rises with dose along a saturating curve, and the simplest description is the occupancy equation: occupancy equals the dose divided by the sum of the dose and a constant, ED50, the dose that occupies half the receptors. Write occupancy as a percentage, O = 100 × D / (D + ED50). Take an antipsychotic with an ED50 of 2 milligrams. At a dose of 1 milligram, O = 100 × 1 / (1 + 2), which is 100 × 1/3, or about 33 percent of D2 receptors occupied. Double the dose to 2 milligrams and O = 100 × 2 / (2 + 2) = 100 × 1/2 = 50 percent, occupancy equal to the dose because the dose now equals the ED50. At 4 milligrams, O = 100 × 4 / (4 + 2) = 100 × 4/6, about 67 percent; at 8 milligrams, O = 100 × 8 / (8 + 2) = 80 percent. Notice the curve's shape: the first doublings buy large gains in occupancy, but as the dose climbs the curve flattens, so pushing from 67 to 80 percent took a further doubling and the next 10 points would take more still. This diminishing return matters clinically because the therapeutic effect of antipsychotics appears in a window of roughly two-thirds to four-fifths D2 occupancy, above which extra dose mainly adds side effects without added benefit. The finding that gives the model its force is that clinical potency tracks receptor affinity across drugs: Creese, Burt, and Snyder showed that the dose of an antipsychotic needed to treat psychosis is predicted, drug for drug, by how tightly it binds the dopamine D2 receptor (Creese et al., 1976). Change the dose and ED50 in the demonstration above and the same equation redraws the occupancy curve.

Discussion

Biological psychiatry has changed what psychiatry is, and it has done so unevenly. On one side stand genuine achievements: a pharmacology that relieves suffering for millions and whose mechanisms are understood at the level of the receptor; a neuroimaging literature that has made the diseased brain visible and testable; and a genomics that has, for the first time, delivered replicable biological findings about disorders that resisted explanation for a century. The dopamine hypothesis, however incomplete, correctly predicted that D2 occupancy would govern antipsychotic action, and the genome-wide studies have turned heritability from an abstraction into specific, synapse-centered biology. On the other side stands a persistent gap between this biology and the clinic. No psychiatric diagnosis is yet made by a biological test; no polygenic score is yet used to guide an individual's treatment; the monoamine hypotheses, though they launched the field, are now seen as at best partial accounts of disorders that are really matters of circuits and development. The field's honest self-assessment is that it has explained a great deal about mechanism while changing everyday diagnosis remarkably little. Its central intellectual problem remains the one Hyman named: the categories inherited from descriptive psychiatry do not carve the brain at its joints, and a biology organized around them keeps failing to cohere. Whether the solution is to redraw the categories around biological dimensions, as the Research Domain Criteria attempt, or to accept that psychiatric kinds are irreducibly defined at the psychological level, is unsettled. What is clear is that the reductive program cannot be abandoned, because it works where it can be applied, and cannot be completed, because a symptom is a fact about a person and not only about a synapse.

Current Directions

The most active front is the effort to convert the genomic haul into biology and, eventually, into treatment. Having established that psychiatric disorders are polygenic and partly shared, the field has turned to functional genomics, asking what the implicated variants actually do, in which cell types and developmental windows they act, and how they converge on the synaptic and neurodevelopmental pathways the schizophrenia data now highlight (Trubetskoy et al., 2022). Patrick Sullivan and Daniel Geschwind have laid out the program explicitly: to define the genetic, genomic, cellular, and diagnostic architectures of psychiatric disorders in an integrated way, moving from a list of associated loci to the neurobiology that connects them to illness (Sullivan & Geschwind, 2019). A parallel agenda has taken stock of psychiatric genomics as a maturing discipline and set its priorities: larger and more ancestrally diverse samples, since polygenic scores trained on people of European descent transfer poorly to others; better integration of rare and common variation; and a realistic account of how far genetic prediction can go in a field where environment matters at every level (Sullivan et al., 2018). Alongside the genetics, the dimensional reframing of diagnosis continues to reshape research funding and design, pulling psychiatric questions back toward the measurable processes of cognition and emotion that the behavioral sciences and the wider study of the brain already investigate. The throughline is a field that has stopped debating whether mental illness is biological and started doing the far harder work of specifying exactly how.

Common Misconceptions

A chemical imbalance is the established cause of depression and other disorders.

The monoamine hypotheses were fruitful working theories, not proven mechanisms, and the evidence never supported a simple deficiency account; depleting monoamines does not reliably cause depression, and disorders are now understood as matters of circuits and development, not a single errant chemical (Schildkraut, 1965).

There is a gene for schizophrenia or for any psychiatric disorder.

There is not. Psychiatric risk is highly polygenic, spread across thousands of common variants of nearly negligible individual effect and partly shared across diagnoses, which is why single-gene and candidate-gene searches failed and genome-wide methods were needed (Sullivan et al., 2012).

Psychiatric diagnoses are validated biological diseases like diabetes.

No psychiatric diagnosis is yet defined or confirmed by a biological test. The DSM categories are defined by symptoms, and treating them as natural biological kinds may itself have hindered the search for their mechanisms (Hyman, 2010).

Glossary

Antipsychotic.

A class of drug that relieves psychosis, chiefly by blocking dopamine D2 receptors; the observation that it did so seeded the dopamine hypothesis.

Biological psychiatry.

The approach to mental illness that locates its causes in brain and body function and pursues its treatment by physical, chiefly pharmacological, means.

Candidate gene.

A gene chosen in advance on biological grounds and tested for association with a disorder; an approach that largely failed to replicate in psychiatry.

Catecholamine hypothesis.

The proposal that depression reflects a functional deficiency, and mania an excess, of catecholamine neurotransmitters such as norepinephrine at key synapses.

Dopamine hypothesis.

The theory that psychosis arises from excess dopamine signaling, inferred from the fact that antipsychotic drugs block dopamine receptors.

Endophenotype.

A heritable, measurable brain-level trait lying between a risk gene and a clinical diagnosis, simpler and nearer the genes than the diagnosis itself.

Genome-wide association study.

A method that scans millions of DNA variants across the whole genome in large samples to find those statistically associated with a disorder, without prior assumptions.

Heritability.

The proportion of variation in a trait or disorder, within a population, attributable to genetic differences among individuals; high for the major psychiatric disorders.

Liability-threshold model.

A model in which many small additive influences sum to a continuous underlying liability, and disorder appears only in those whose liability exceeds a threshold.

Monoamine.

A class of neurotransmitter that includes dopamine, norepinephrine, and serotonin; the target of the drugs behind psychiatry's first biological theories.

Neuroimaging.

Techniques that visualize the structure or activity of the living brain, used to link psychiatric symptoms to abnormalities of specific neural circuits.

Neurotransmitter.

A signaling molecule released by a neuron across the synapse to excite or inhibit the next cell; the physical currency of the brain's communication.

Polygenic score.

A single number summing the disorder-associated effects of many genetic variants across a person's genome to estimate genetic liability.

Psychopharmacology.

The study of how drugs act on the mind through their effects on the brain; the applied science born of the monoamine hypotheses.

Receptor occupancy.

The fraction of a given receptor population bound by a drug at a given dose; for antipsychotics it tracks clinical effect within a therapeutic window.

Reductionism.

The explanatory strategy of accounting for a complex phenomenon by the properties of its parts; here, explaining a syndrome by circuits, synapses, and genes.

Reification.

The error of treating an abstract descriptive category, such as a DSM diagnosis, as though it were a concrete natural entity with a fixed biological essence.

Research Domain Criteria.

A research framework that sets diagnostic categories aside to study dimensions of function, such as reward and fear, across levels from genes to behavior.

Key Researchers

Nancy C. Andreasen. Neuropsychiatrist at the University of Iowa and a founder of the neuroimaging study of schizophrenia; she argued for a scientific psychopathology linking symptoms to measurable brain abnormalities and long edited the American Journal of Psychiatry. ORCID - Wikipedia - Wikidata

Arvid Carlsson (1923-2018). Swedish pharmacologist and Nobel laureate; he established dopamine as a neurotransmitter and showed that antipsychotics act on dopamine signaling, supplying the mechanism behind the dopamine hypothesis. Wikipedia - Wikidata

Steven E. Hyman. Neuroscientist and psychiatrist at Harvard and the Broad Institute and a former director of the National Institute of Mental Health; he argued that the reification of DSM categories had impeded the search for their biology. ORCID - Wikipedia - Wikidata

Thomas R. Insel. Psychiatrist and neuroscientist and a former director of the National Institute of Mental Health; he championed the Research Domain Criteria as a biologically grounded alternative to symptom-based diagnosis. ORCID - Wikipedia - Wikidata

Eric R. Kandel. Neuroscientist at Columbia University and Nobel laureate; his call for a new intellectual framework for psychiatry, in which all mental processes are brain processes, became the field's modern charter. ORCID - Wikipedia - Wikidata

Emil Kraepelin (1856-1926). German psychiatrist and the founder of modern psychiatric nosology; his insistence that mental illnesses are natural biological disease entities framed the classification psychiatry still uses. Wikipedia - Wikidata

Michael C. O'Donovan. Psychiatric geneticist at Cardiff University and a leader of the Psychiatric Genomics Consortium; he was a senior author of the genome-wide studies that mapped schizophrenia risk to synaptic biology. ORCID - Wikipedia - Wikidata

Joseph J. Schildkraut (1934-2006). American psychiatrist at Harvard; his catecholamine hypothesis of affective disorders synthesized the pharmacology of mood into psychiatry's most influential early biological theory of depression. Obituary

Solomon H. Snyder. Neuroscientist at Johns Hopkins University; with Creese and Burt he showed that antipsychotic clinical potency is predicted by affinity for the dopamine D2 receptor, tying a molecule directly to a syndrome. Wikipedia - Wikidata

Patrick F. Sullivan. Psychiatric geneticist at the University of North Carolina and the Karolinska Institutet and a founder of the Psychiatric Genomics Consortium; his work defined the polygenic architecture and the research agenda of the field. ORCID - Wikipedia - Wikidata

Daniel R. Weinberger. Psychiatrist and neuroscientist at the Lieber Institute and Johns Hopkins; with Meyer-Lindenberg he developed the intermediate-phenotype approach linking genes to brain function through imaging. ORCID - Wikipedia - Wikidata

Frequently Asked Questions

What is biological psychiatry? Biological psychiatry is the approach to mental illness that treats disorders as disturbances of brain function and seeks their causes and treatments in the brain's chemistry, circuitry, and genetics. It does not deny that psychological and social factors matter, but it insists that the biological level is real, measurable, and essential to explaining and treating psychiatric illness (Kandel, 1998).

What is the dopamine hypothesis of schizophrenia? It is the theory that psychosis involves excess dopamine signaling in the brain. The idea was inferred from pharmacology: antipsychotic drugs were found to block dopamine receptors, and their ability to relieve psychosis suggested that overactive dopamine transmission was part of its mechanism (Carlsson & Lindqvist, 1963).

Is depression caused by a chemical imbalance? The chemical-imbalance idea comes from the catecholamine hypothesis, an influential early theory that depression reflects a deficiency of neurotransmitters such as norepinephrine. It was a productive working model but was never proven, and depression is now understood as a disorder of brain circuits and development rather than a single missing chemical (Schildkraut, 1965).

How do antipsychotic drugs work? Antipsychotics act mainly by occupying dopamine D2 receptors, and their clinical potency across drugs is predicted by how tightly they bind that receptor. Clinical benefit appears within a window of roughly two-thirds to four-fifths receptor occupancy, above which extra dose mostly adds side effects (Creese et al., 1976).

Are psychiatric disorders genetic? The major psychiatric disorders are substantially heritable, but no single gene causes them. Risk is highly polygenic, spread across thousands of common genetic variants of tiny individual effect, and a good deal of that risk is shared across different diagnoses (Sullivan et al., 2012).

Is there a blood test or brain scan that diagnoses mental illness? No. Despite decades of neuroimaging and genetics, no psychiatric diagnosis is yet made by a biological test. Diagnoses remain defined by symptoms, and finding reliable biological markers is one of the field's central unmet goals (Andreasen, 1997).

What is the Research Domain Criteria framework? The Research Domain Criteria, or RDoC, is a research framework from the National Institute of Mental Health that sets aside traditional diagnostic categories and instead studies dimensions of function, such as reward and fear, across levels from genes to circuits to behavior, on the premise that biology respects these dimensions rather than the diagnoses (Insel et al., 2010).

Why do biological psychiatry's findings so rarely change everyday treatment? Because a large gap remains between mechanism and the clinic. The field has learned a great deal about receptors, circuits, and genes, but psychiatric categories are still defined by symptoms, no polygenic score yet guides an individual's care, and the biology has not mapped neatly onto the diagnoses clinicians use (Hyman, 2010).

References