The “Primacy of Mania” Hypothesis (PM Hypothesis)

On the proposition that depression is a sequela of mania

(CL) — Shinagawa Psychosomatic Clinic, Free Notes Vol. 5 2026.03.05

I. Introduction — What This Hypothesis Is Asking

In psychiatry, depression has long been treated as a ‘primary condition.’ Even within the DSM system, Major Depressive Disorder (MDD) is defined as a disease entity independent of Bipolar Disorder, and clinically, the premise that ‘depression comes first’ has been the starting point of treatment.

The PM Hypothesis (Primacy of Mania hypothesis) fundamentally challenges this premise. The core proposition of this hypothesis can be formulated as follows:

Behind every instance of depression, there necessarily exists a prior state of cerebral excitation — not necessarily a clear manic or hypomanic episode, but at minimum a phase of neurobiological hyperactivation. Depression is not a primary disease; it is a sequel to an excitation phase.

This is not merely a question of the sequence of phases. If this hypothesis is correct, the entire paradigm of ‘depression treatment’ is called into question: the validity of antidepressants as the first-line choice, the independence of ‘unipolar depression’ as a disease entity, and ultimately the reorganization of the psychiatric diagnostic system itself.

II. The Proponents and Theoretical Background

2-1. S. Nassir Ghaemi and His Position

The most systematic proponent of the PM Hypothesis is S. Nassir Ghaemi (Tufts University School of Medicine, Department of Psychiatry; also spelled ‘Ghamie’). He developed this hypothesis in the context of bipolar spectrum research.

Ghaemi’s position can be summarized in one sentence: ‘Unipolar depression is overdiagnosed, and a substantial portion of clinical depression is the depressive phase of undiagnosed bipolar disorder.’ This connects with Hagop Akiskal’s Bipolar Spectrum hypothesis, but the PM Hypothesis goes further, asserting a neurobiological mechanism that places the manic/excitation phase as the origin of the pathological process.

Major works:

• Mood Disorders: A Guide to Living with Depression and Manic-Depression (2007)
• A First-Rate Madness: Uncovering the Links Between Leadership and Mental Illness (2011)
• Numerous review articles in Bipolar Disorders, Acta Psychiatrica Scandinavica, and other journals

2-2. The Intellectual Lineage of the Theory

The PM Hypothesis did not emerge as a mutation but can be understood as the convergence of several intellectual currents:

(a) Kraepelin’s theory of manic-depressive illness Kraepelin (1913) understood manic-depressive illness as a single pathological entity (manic-depressive insanity), viewing mania and depression as different poles of the same disease process. Based on his observations, in many patients manic states precede or intermingle with depression.

(b) The inversion of the catecholamine hypothesis Schildkraut’s classical catecholamine hypothesis (1965) understood depression as a deficiency of noradrenaline and dopamine, but the PM Hypothesis focuses on the prior phase — the phase of excess release. This is a dynamic model of excess → depletion.

(c) Clinical observations of post-manic depression A depressive state following a manic episode (post-manic depression) has long been recognized in bipolar disorder. Ghaemi and colleagues extend this phenomenon by interpreting it not as peculiar to bipolar disorder, but as the manifestation of a more universal mechanism.

(d) The Kindling model Post’s (1992) kindling hypothesis holds that repeated episodes sensitize the brain so that it progressively generates episodes without external stimulation. Combined with the PM Hypothesis, this constructs a progression model of ‘mania → depression → further amplified mania → deeper depression.’

III. The Content of the Hypothesis — A Detailed Explanation

3-1. How ‘Mania’ Is Defined

The ‘manic state’ in the PM Hypothesis is not limited to the full Manic Episode as defined by DSM. This is the core of the theory and also the focal point of criticism.

The hierarchy of preceding excitation states envisioned by Ghaemi and colleagues:

Level	Clinical Equivalent	Characteristics
Full mania	DSM Manic Episode	Requires inpatient care; may involve psychotic symptoms
Hypomania	DSM Hypomanic Episode	Minimal functional impairment; patient feels well
Subsyndromal hypomania	Subsyndromal hypomania	Below diagnostic threshold; increased energy, reduced sleep, talkativeness
Neurobiological hyperactivation	Clinically subclinical	Sleep changes, thermoregulatory abnormalities, HPA axis activation

This expanded definition simultaneously broadens the theory’s scope and makes verification difficult. ‘Neurobiological hyperactivation’ is clinically invisible and in many cases can only be confirmed retrospectively.

3-2. Proposed Neurobiological Mechanisms

The pathophysiological model presented by the PM Hypothesis can be organized as follows:

(Step 1) Precipitants of neural excitation

• Sleep deprivation (social factors, circadian rhythm disturbances)
• HPA axis and sympathetic nervous system hyperactivation due to stress responses
• Seasonal changes (changes in light exposure)
• Substance use (caffeine, alcohol, various psychoactive substances)

(Step 2) Neurochemical changes during the excitation phase

• Dopamine system hyperactivation: reinforcement of reward circuits, increased risk-taking behavior
• Excess noradrenaline release: elevated arousal and activity
• Serotonin system fluctuations
• Glutamate system activation: increased excitatory neurotransmission
• Transient elevation of brain-derived neurotrophic factor (BDNF)

(Step 3) Depletion and downregulation As a protective and compensatory mechanism following hyperactivation:

• Receptor downregulation
• Relative depletion of neurotransmitters
• HPA axis exhaustion (cortisol dysregulation)
• Delay and distortion of circadian rhythms
• Subsequent neuroinflammation (glial activation)

(Step 4) Onset of depressive state The depletion and dysregulation described above trigger depressive symptoms. Only at this stage does the patient become clinically recognized as experiencing ‘depression.’

What is decisive in this dynamic model is that by the time the clinician identifies ‘depression,’ the preceding phase (the excitation phase) has already ended, or the patient themselves has not recognized it as a symptom.

3-3. Relationship to Sleep Deprivation Therapy

As a powerful piece of indirect evidence for the PM Hypothesis, the antidepressant effect of therapeutic sleep deprivation (sleep deprivation therapy / wake therapy) is cited.

Sleep deprivation induces a mild manic-like state in both healthy individuals and patients, and often leads to a dramatic improvement in mood the following day. Proponents of the PM Hypothesis interpret this as follows:

In a depressive state, the excitation → depression chain has already been completed. Sleep deprivation artificially introduces a new excitation phase, temporarily pulling the patient away from depression. However, the well-known clinical fact that depression rebounds after recovery sleep occurs because the excitation → depletion cycle is set into motion again.

This interpretation is not direct proof, but it supports the coherence of the mechanism.

3-4. The Connection to the Problem of Antidepressants

What Ghaemi emphasizes in particular is the possibility that the administration of antidepressants to patients diagnosed with unipolar depression can induce a switch to mania.

The logic that ‘if the patient switches to mania on antidepressants, then that patient must have had bipolar disorder (Bipolar II) to begin with’ is widely accepted in psychiatric practice, but the PM Hypothesis takes this a step further:

• Manic switching caused by antidepressants can be interpreted not only as ‘the manifestation of a pre-existing bipolar diathesis’ but also as ‘the pharmacologically forced induction of an excitation phase.’
• The observation that antidepressants worsen depression in the long term (Post’s switching hypothesis) is consistent with a model in which the drug accelerates the excitation → depression cycle.
• The interpretation is also possible that ‘emotional blunting’ caused by SSRIs prevents escape from depression by suppressing the excitation phase.

3-5. Diagnostic Implications

If the PM Hypothesis is correct, the following diagnostic reconsiderations are required:

1. Much of ‘unipolar depression’ is latent bipolar disorder: hypomania is being missed, or the patient is recognizing it as a period of good health.
2. The life history prior to the first onset of depression should be evaluated in detail: retrospective evaluation of subsyndromal hypomanic signs such as changes in sleep, activity level, talkativeness, hyperactivity, financial extravagance, and increased sexual activity.
3. Reconsideration of antidepressant monotherapy: priority use of mood stabilizers and atypical antipsychotics.

IV. Supporting Evidence and Arguments

The evidence supporting the PM Hypothesis is organized below according to strength.

4-1. Strong Support

(a) High frequency of the mania → depression sequence in bipolar disorder In longitudinal studies of Bipolar I and II disorder, the frequency of depressive episodes following manic or hypomanic episodes is high (particularly in Bipolar I, with reports of 50–70% transitioning from mania to post-manic depression). This can be understood as a special case of ‘mania preceding all depression,’ though it remains a finding limited to the bipolar patient group.

(b) Biological specificity of post-manic depression Post-manic depression is suggested to differ from spontaneous depression in its timing, treatment responsiveness, and neuroimaging findings, supporting the neurobiological specificity of ‘secondary depression.’

(c) Preventive effect of mood stabilizers on depression The fact that mood stabilizers such as lithium and lamotrigine prevent depressive episodes is consistent with a model in which suppressing manic states leads to the suppression of depression.

4-2. Indirect Support

(a) Circadian rhythm research Many patients with depression show circadian rhythm disturbances (phase advance or delay), which may suggest that shortened sleep and sleep changes during an excitation phase preceded the depression.

(b) Sleep changes prior to first-onset depression Prospective studies report sleep disturbances (shortened sleep duration, early morning awakening) in the weeks preceding the onset of depression, which can be interpreted as signs of subsyndromal excitation.

(c) Underdiagnosis of bipolar disorder Large-scale epidemiological studies (e.g., Angst et al., 2003) suggest that with rigorous assessment, the prevalence of bipolar spectrum disorders reaches 5–10% of the population, and that the diagnosis of unipolar depression contains a considerable number of latent bipolar patients.

(d) Dynamics of the HPA axis Acute stress (short-term cortisol elevation = a proxy measure for the excitation phase) followed by HPA axis hyporeactivity (observed in depressive states) is consistent with a model of hyperactivation → exhaustion.

V. Criticisms and Counterarguments

The counterarguments to the PM Hypothesis are diverse and powerful. The main ones are detailed below.

5-1. Conceptual and Methodological Criticisms

Criticism 1: The definition of ‘excitation state’ is excessively expanded and unfalsifiable

As long as the PM Hypothesis includes neurobiological hyperactivation as a broad preceding condition, it can be applied retrospectively to virtually any case of depression. Some form of stress response (HPA axis activation) occurs in many people, allowing one to say that ‘there was excitation before the depression.’

Viewed through Karl Popper’s criterion of falsifiability, insofar as this hypothesis includes clinically subclinical neurobiological hyperactivation, it approaches a proposition that is in principle impossible to falsify — always true. This undermines the scientific value of the hypothesis.

Criticism 2: The problem of retrospective bias

The claim that ‘there was always hypomania before depression’ is often based on retrospective evaluation. However:

• Patients remember hypomanic periods as ‘good times’ and do not report them as symptoms.
• If the interviewer assumes bipolarity in their assessment, hypomanic signs are easier to find.
• Screening tools such as the Hypomania Checklist (HCL-32) prioritize sensitivity and thus tend to lead to overdiagnosis.

This epistemological problem is serious and makes it impossible to exclude the possibility that ‘the hypomania did not go undetected — it simply did not exist.’

5-2. Epidemiological and Genetic Criticisms

Criticism 3: The genetic separation of unipolar depression and bipolar disorder

Large-scale genomic studies (PGC: Psychiatric Genomics Consortium) have revealed that while unipolar depression and bipolar disorder share some genomic risk, each also has its own unique genetic contributions.

• There are risk variants specific to MDD (unipolar).
• There are risk variants specific to Bipolar I.
• The two do not form a complete genetic continuum.

This contradicts the strong form of the PM Hypothesis, which holds that all unipolar depression is latently bipolar.

Criticism 4: Inconsistencies in family studies

If all depression were essentially bipolar, then the frequency of bipolar disorder would be elevated in first-degree relatives of depressed patients. However, family aggregation studies show that unipolar depression and bipolar disorder exhibit mutually independent patterns of familial aggregation (Gershon et al., McGuffin et al., etc.).

5-3. Neurobiological Criticisms

Criticism 5: Differences in structural brain changes

Unipolar depression and bipolar disorder differ in many respects with regard to brain structure and function:

• Patterns of prefrontal cortex volume reduction differ.
• Hippocampal volume reduction is more pronounced in unipolar depression, according to some reports.
• The distribution of white matter lesions differs.
• The patterns of network abnormalities on functional MRI differ.

If depression were always a sequela of mania, the neuroanatomical differences between the two conditions would be difficult to explain.

Criticism 6: The existence of animal models

In established animal models of depression, such as the Chronic Mild Stress (CMS) model, the Learned Helplessness model, and the Social Defeat Stress model, depressive-like behavior is induced without a preceding excitation phase.

Experimental animal evidence supports the conclusion that ‘depression can arise primarily without excitation.’

5-4. Clinical Criticisms

Criticism 7: The existence of fully unipolar courses

In long-term longitudinal studies (e.g., Angst et al.’s Zurich Cohort, Judd et al.’s CIDSP study), there certainly exist depressed patients who have never experienced a manic or hypomanic episode throughout their lifetime.

It is possible to hypothesize a ‘neurobiological hyperactivation’ in these patients, but such a hypothesis is inconsistent with the clinical trajectory and constitutes a non-verifiable assumption.

Criticism 8: Mood stabilizers do not always prevent depression

While evidence exists that mood stabilizers prevent depression, their effect is incomplete, and the clear efficacy of mood stabilizers for unipolar depression is not as well established as it is for bipolar disorder.

The prediction that ‘suppressing mania will prevent depression’ has not been fully realized.

Criticism 9: The independence of ‘pure melancholia’

Melancholic depression is neurobiologically distinct (marked HPA axis hyperactivation, morning worsening, a specific pattern of appetite and anhedonia loss), and some researchers support it as a disease entity separate from the bipolar spectrum (Bernard Carroll, Gordon Parker, etc.).

Proponents of melancholia counter the PM Hypothesis with the argument that ‘typical endogenous depression exists even without evidence of a preceding manic state.’

5-5. Criticisms Regarding Therapeutic Implications

Criticism 10: Denial of antidepressant efficacy is premature

Ghaemi is skeptical of the efficacy of IPT (Interpersonal Therapy) and antidepressants for bipolar depression, but while the STEP-BD study (Sachs et al., 2007) showed that SSRIs/bupropion added no benefit for bipolar depression, guidelines such as CANMAT still conditionally support their use.

Furthermore, the efficacy of antidepressants for unipolar depression (demonstrated in numerous RCTs) is an incontrovertible body of evidence, and the claim suggested by the PM Hypothesis — that ‘antidepressants worsen depression in the long term’ — requires stronger prospective evidence.

VI. The Current Position and Evaluation of the Hypothesis

6-1. Distinguishing the ‘Strong Form’ from the ‘Weak Form’

The PM Hypothesis receives different evaluations depending on the strength of its claims:

Strong form: ‘Every depressive state necessarily has a preceding mania or excitation. Unipolar depression does not exist.’ → This runs counter to evidence from genetics, epidemiology, and animal research, and is difficult to support.

Weak form: ‘At least the depressive phase of bipolar disorder should be understood as a sequel to the excitation phase, and the bipolar spectrum is potentially far wider than previously assumed, possibly encompassing many cases of unipolar depression.’ This carries clinically important implications and can be partially supported.

In psychiatric clinical practice, the implications of the weak form — namely, ‘do not miss the hypomania that precedes depression’ and ‘be cautious about antidepressant monotherapy’ — hold practical value.

6-2. The Merits of the Theory

The merits of the PM Hypothesis should be evaluated beyond the binary of ‘whether it is right or wrong.’

4. It sharply identified the problem of underdiagnosis of bipolar disorder: this has actually changed clinical practice.
5. It served as a warning against the careless prescription of antidepressants in monotherapy: useful in preventing actual therapeutic harm.
6. It provided a perspective for understanding depression not as a static state but as a dynamic process: enriching the framework of depression research.
7. It theoretically supported the impulse toward a return to Kraepelinian unitary psychosis theory: a critique of the artificiality of diagnostic boundaries.

6-3. The Limitations of the Theory

8. The problem of falsifiability remains unresolved.
9. The operational definition of neurobiological hyperactivation remains vague.
10. The strong form cannot be maintained: the independence of disease entities shown by genetics cannot be ignored.
11. The risk of excessive clinical application: the risk of a slide toward the oversimplification that ‘all depression should be treated with mood stabilizers.’

VII. A Perspective as a Clinical Psychiatrist — A Structural Reading

Finally, I would like to append an attempt to read this hypothesis at one level of abstraction higher.

What the PM Hypothesis fundamentally asks is the question of the temporal structure of the phenomenon of depression. It demands that depression be understood not as a synchronic state but as one phase of a diachronic process.

This also connects to the context of existentialist psychiatry. Just as Binswanger, in his ‘Daseinsanalyse,’ described depression as a ‘disturbance of temporality,’ depression presents itself as ‘suffering in the here and now,’ but its formation is preceded by a temporal context. The PM Hypothesis can also be read as an attempt to articulate this temporal precedence in the language of neurobiology.

Or, in the framework of predictive processing theory: a manic state is a phase of increased prediction error and upward revision; depression arises as overcorrection — a ‘collapse of precision.’ This reading enables a phenomenological-neuroscientific interpretation consistent with the PM Hypothesis.

From a structural standpoint, however, the PM Hypothesis already contains its answer in the very framing of the question of the ‘primary versus secondary nature of depression,’ carrying the danger of question-begging. What should truly be asked may not be ‘what is primary and what is secondary,’ but ‘whether the primary/secondary distinction is productive for understanding pathology.’

When that question is reached, the PM Hypothesis will be given its own place in the history of psychiatric concepts as ‘an attempt to theorize a single powerful intuition.’

References (Representative Works)

Ghaemi SN. Mood Disorders: A Guide to Living with Depression and Manic-Depression. Johns Hopkins University Press, 2007.

Ghaemi SN, et al. “Diagnosing bipolar disorder and the effect of antidepressants: a naturalistic study.” Journal of Clinical Psychiatry, 2000.

Akiskal HS, Pinto O. “The evolving bipolar spectrum.” Psychiatric Clinics of North America, 1999.

Post RM. “Transduction of psychosocial stress into the neurobiology of recurrent affective disorder.” American Journal of Psychiatry, 1992.

Sachs GS, et al. “Effectiveness of adjunctive antidepressant treatment for bipolar depression.” NEJM, 2007.

Angst J, et al. “Toward a re-definition of subthreshold bipolarity.” Journal of Affective Disorders, 2003.

Carroll BJ. “Psychopathology and neurobiology of manic-depressive disorders.” In: Psychopathology and the Brain, 1991.