Learning outcomes
By the end of this lesson you should be able to: state the biopsychosocial model and explain what it adds to a biological account; explain the stress-vulnerability (diathesis-stress) framework and how it generates testable predictions; describe the evidence from twin, adoption and genome-wide association studies and what heritability does and does not mean; outline the major neurobiological hypotheses for depression, psychosis and anxiety; describe the psychological and social contributions, including conditioning, cognitive schema, attachment, life events, poverty and social defeat; and construct a multi-level formulation for a patient rather than reducing them to a diagnosis.
Prerequisites
Lesson 5 (Classification and assessment). You need the idea of a syndrome, and the distinction between factors that predispose, precipitate and perpetuate, before the causal claims in this lesson can be read carefully.
The big idea
Engel's argument was not that biology does not matter — he was himself a consultant physician and internist — but that a model which explains everything at the cellular level explains nothing at the level at which the patient lives. The practical pay-off is that the same patient can be treated at several levels at once: a mood stabiliser such as lithium or valproate for a manic episode, cognitive therapy for the depressive thinking that follows it, and a debt adviser for the spending that precipitated it. Each intervention targets a different level, and none substitutes for another.
An easy way to remember the levels is to ask: what is this person made of (biological), what do they do and believe (psychological), and where in the world are they situated (social)? A single-factor account — 'a chemical imbalance', 'a bad upbringing', 'a chemical in the water' — should always be read as a hypothesis of last resort, not an explanation.
The stress-vulnerability framework
Zubin and Spring's vulnerability model (1977) separates a stable disposition — the diathesis — from transient stressors, and asks when the two coincide. The clinical value is twofold: it explains why patients relapse when stressed and improve when settled, and it directly generates intervention. If a patient has a high vulnerability, the highest-yield interventions may be environmental — reducing exposure to high expressed emotion, stabilising sleep, avoiding cannabis — rather than merely adding medication.
| Phase | Meaning | Examples | What the clinician does about it |
|---|---|---|---|
| Predisposing | Factors present from birth or early life that raise vulnerability | Genetic loading, obstetric complications, childhood maltreatment, developmental intellectual disability | Informs prognosis and family counselling; rarely modifiable but shapes prevention |
| Precipitating | Events that trigger the current episode | Bereavement, childbirth (postpartum psychosis), a first psychotic episode after cannabis use, a job loss | The focus of immediate intervention and of crisis planning |
| Perpetuating | Factors that keep the episode going once it has started | Avoidance behaviour, chronic insomnia, high expressed emotion in the family, secondary gain, substance misuse, untreated comorbidity | Where most therapeutic gain is available — the part that responds to treatment |
| Protective | Factors that lower the probability and severity of relapse | Social support, employment, adherence, insight, secure housing, stable relationships | Deliberately cultivated; they change the course more than most acute interventions |
NoteA formulation is the biopsychosocial model applied to one patient: predisposing, precipitating, perpetuating and protective factors across all three levels, leading to an explicit management plan. The diagnosis says what the disorder is; the formulation says why this person has it now, and what to do next.
Genetic contributions
The classic evidence rests on comparing people who differ in genetic and environmental sharing. Twin studies show higher concordance in monozygotic than in dizygotic pairs, and adoption studies separate genetic relatedness from rearing environment. In schizophrenia, monozygotic concordance of roughly 40 to 50 percent and dizygotic of roughly 10 to 17 percent supports strong genetic loading, but the monozygotic figure being well under 100 percent establishes that genes are not sufficient. Genome-wide association studies have since identified hundreds of common variants of small individual effect, and copy number variants (such as 22q11.2 deletion) of larger effect.
| Disorder | Approximate heritability | What the pattern of inheritance implies | Known environmental contributions |
|---|---|---|---|
| Schizophrenia | Around 60-80 percent in twin studies | Strongly polygenic; thousands of variants each of small effect; no single gene causes it | Urban birth and upbringing, migration and minority status, childhood adversity, cannabis use in adolescence, obstetric complications |
| Bipolar disorder | Around 60-85 percent | Even more strongly genetic; overlapping liability with schizophrenia in GWAS data | Sleep deprivation, childbirth, life events, antidepressant switching |
| Major depressive disorder | Around 35-40 percent | Moderate heritability for a broad liability; gene-environment interaction is prominent | Loss events, chronic strain, poverty, isolation, medical illness |
| Anxiety disorders | Around 30-50 percent | Shared liability across the anxiety disorders; trait neuroticism sits upstream | Adverse parenting, overprotective caregiving, traumatic events |
| Alcohol dependence | Around 50-60 percent | Substantially genetic via impulsivity and alcohol metabolism; ALDH2 variant strongly protective in East Asian populations | Availability, price, social norms, trauma, peer use |
| Autism spectrum disorder | Around 60-90 percent | Highly polygenic, large contribution from de novo variants; substantial overlap with intellectual disability | No established environmental cause; the MMR-vaccine link has been retracted and discredited |
Watch outHeritability is a statement about variance in a population, not about a person and not about destiny. Two patients with identical diagnoses can have entirely different genetic loading, and identical twins who are discordant for schizophrenia demonstrate that the non-genetic half is real. Quoting heritability to a worried family should be paired with the more useful statement: what raises risk mostly is what treatment can help with.
Gene-environment interplay takes two forms. In gene-environment correlation, genes shape exposure — a genetically inhibited child seeks fewer novel encounters, and a genetically impulsive adolescent seeks more risky peers. In gene-environment interaction, the effect of an environmental exposure depends on genotype: the best-known psychiatric example is that the effect of stressful life events on depressive episodes is stronger in people with particular variants in stress-response pathways, though the effect sizes are small and replication has been uneven.
Neurobiology
| Hypothesis | System | Principal evidence | Status of the claim |
|---|---|---|---|
| Dopamine hypothesis of psychosis | Mesolimbic dopamine signalling; striatal D2 receptor availability | Amphetamines and dopamine agonists induce psychosis and worsen schizophrenia; antipsychotics are D2 antagonists; imaging shows elevated striatal dopamine synthesis capacity | The best-supported neurochemical account of positive symptoms, but it is one part of a wider story and does not explain negative symptoms |
| Serotonergic hypothesis | Serotonergic transmission and receptor systems | SSRIs increase synaptic serotonin and relieve depression; triptans and hallucinogens act on serotonin receptors | Seriously limited as a causal account: depression onset is delayed 2-4 weeks after serotonin rises, and SSRIs relieve depression without correcting any measurable deficit |
| Monoamine hypothesis of depression | Serotonin, noradrenaline and dopamine | Reserpine depletes monoamines and can induce depression; most antidepressants act on monoamine reuptake or metabolism | Superseded as a complete account — it describes drug targets, not cause; neurotrophic and stress-axis models now sit alongside |
| Glutamate hypothesis of psychosis and depression | NMDA receptor hypofunction and glutamate-glutamine cycling | Ketamine and phencyclidine produce schizophreniform symptoms and rapid antidepressant effects; postmortem and spectroscopy findings are suggestive | The most active current line for both depression and negative symptoms; esketamine is derived from it |
| HPA axis dysregulation | Hypothalamic-pituitary-adrenal axis, CRH and cortisol | Non-suppression on dexamethasone suppression testing in a proportion of patients with melancholic depression; elevated cortisol in chronic stress | A state marker of severity and of chronic stress, not a diagnostic test |
| Neurotrophic and inflammatory | BDNF, neurogenesis; cytokine signalling | BDNF rises with effective antidepressant treatment; inflammation is associated with depression in chronic illness | Moderate; relevant to illness behaviours, fatigue and sickness behaviour in medical depression |
Watch out'A chemical imbalance' is not a psychiatric explanation. It is a drug-advertising simplification, it is not measurable in the clinic, and it leads patients into the belief that tablets alone will fix a complex problem. Depression is not a serotonin deficiency; SSRIs happen to act on serotonin and take weeks to change mood. Explain treatment as 'this medication corrects a pattern of brain signalling involved in your symptoms', and be honest that the mechanism is not fully understood.
Structural and functional imaging contributes context rather than diagnosis: patients with schizophrenia show slightly enlarged ventricles and reduced hippocampal and prefrontal cortical volumes on group averages, PTSD shows smaller hippocampal volume, and mood disorders show altered amygdala and prefrontal activation. None of these is diagnostic for an individual, and no psychiatric diagnosis is made from a scan. Neuropsychological profiles are more useful at the bedside: the frontal-subcortical pattern in Parkinson's disease and Huntington's disease, the episodic-dominant profile in early Alzheimer's disease, and the patchy pattern in vascular dementia.
Psychological contributions
| Mechanism | Theory | How it produces disorder | Therapy implied |
|---|---|---|---|
| Classical conditioning | Watson and Rayner; Mowrer two-factor | Neutral cues acquire fear by pairing with an aversive event and the response is maintained by avoidance | Graded exposure and response prevention |
| Operant conditioning | Skinner | Depression as a loss of environmental reinforcement — activities stop paying, so they stop happening, so they pay less | Behavioural activation: scheduling valued, rewarding activity |
| Cognitive schemas and the triad | Beck | Negative views of self, world and future; dysfunctional assumptions activated by matching events | Cognitive restructuring; cognitive therapy for relapse prevention |
| Learned helplessness | Seligman | Repeated exposure to uncontrollable outcomes produces the belief that action is pointless, plus the cognitive attributional style that makes it general | Attribution retraining; problem-solving therapy |
| Attachment and loss | Bowlby | Early loss or insecure attachment shapes the internal working model of relationships, predicting later vulnerability to depression | Psychoanalytic and attachment-based therapies |
| Psychodynamic conflict and defence | Freud and successors | Unacceptable impulses and affects are managed by defences that distort reality and symptoms | Psychodynamic psychotherapy |
| Sexual conditioning | Conditioning of arousal patterns | Unusual stimuli acquire erotic significance, and the pattern is reinforced by orgasm | Behavioural retraining where the pattern is distressing or harmful |
Social contributions
The social environment is not background noise; it is causal. Brown and Harris established that depressive episodes in women followed loss and humiliation events, that the effect was amplified by the absence of a confiding relationship, and that working-class women with children at home had four times the rate of depression of middle-class women without. Social factors also determine whether a disorder is recognized, treated and supported.
| Social factor | Effect | Disorder it most clearly affects |
|---|---|---|
| Life events and chronic strain | Precipitates depressive episodes; humiliation and entrapment events are particularly potent | Depression, PTSD, adjustment disorders |
| Poverty, unemployment and debt | Prevalence and outcome worsen; treatment access falls; debt adds a chronic, unremitting stressor | Depression, anxiety, alcohol and drug dependence, self-harm |
| Loneliness and isolation | Loneliness predicts depression, anxiety and dementia, and carries mortality risk comparable with smoking | Depression, Alzheimer's disease, self-harm in older people |
| Urban birth and upbringing | Dose-related increase in risk for schizophrenia, independent of family history | Schizophrenia spectrum disorders |
| Migration, minority status and discrimination | Elevated rates of psychosis; misdiagnosis is also more frequent | Schizophrenia and affective psychosis |
| High expressed emotion in the family | Criticism, hostility and emotional overinvolvement predict relapse two to three times over | Schizophrenia, bipolar disorder, anorexia nervosa |
| Adverse childhood experiences | Dose-related increase in depression, PTSD, substance misuse and self-harm | Nearly all psychiatric disorders |
| Stigma | Reduces help-seeking and treatment adherence, worsens employment outcomes, and adds an internalised stressor | All psychiatric disorders |
NoteStigma is both a cause and a consequence. Public stigma (stereotypes held by others), self stigma (beliefs a patient holds about themselves) and structural stigma (legal and institutional arrangements) all produce measurable clinical harm. Language is a site of this: 'schizophrenic' as a noun, 'addict' as a synonym for a person, and 'a danger to the public' as a description of psychosis all move practice in the wrong direction.
Common mistakes
Watch outExplaining a disorder at one level and calling it explained. 'It is genes' does not tell you which life event triggered this relapse; 'it is trauma' does not tell you why one traumatised person becomes depressed and another does not. Aetiology is multi-level, and the intervention usually follows the level at which the mechanism is available.
Watch outInferring causation from a correlation in epidemiology. Schizophrenia and cannabis use co-occur, but most cannabis users never develop psychosis and most patients have never used it — and reverse causation (prodromal patients self-medicate) has been demonstrated in cohorts. Associations constrain hypotheses; they do not settle them.
Watch outWriting 'psychosocial' as a euphemism for 'not really organic'. Psychosocial factors are causal and objective — unemployment, a bereavement, a hostile relative — and 'organic' is not a synonym for real. Health psychology has repeatedly found that social position alters measured biological risk.
Check your understanding
A patient has a father with schizophrenia, used cannabis heavily at seventeen, was born in a city and grew up there, has lost several jobs, and has a highly critical mother. Map these to the four phases of the stress-vulnerability model.
Why is it misleading to tell a patient with depression that they have 'a chemical imbalance'?
How does the two-factor theory of avoidance, from Lesson 2, fit into the stress-vulnerability model?
Summary and next
Aetiology is multi-level: genetic liability is real and overwhelmingly polygenic, neurobiology explains parts of syndromes rather than whole ones, psychological mechanisms explain how symptoms are acquired and maintained, and social factors carry independent causal weight. The stress-vulnerability framework organises these into predisposing, precipitating, perpetuating and protective factors, and it converts directly into a management plan.
Next: the psychotic disorders. You will meet delusions, hallucinations and thought disorder as phenomena, and see how the five As of schizophrenia, the differential diagnosis of psychosis, and the management of antipsychotic side effects connect back to the dopamine hypothesis you have just read.
Multiple choice
Which finding most directly demonstrates that the stress-vulnerability model adds predictive power beyond vulnerability alone?
Which statement about heritability of schizophrenia is correct?
Which of these is the strongest current neurochemical account of the positive symptoms of schizophrenia?
A patient's depressive episode began after redundancy and homelessness, and persists because of debt. Which level does this aetiology best illustrate?
Which statement correctly describes the role of cannabis in schizophrenia risk?