Longitudinal Formulation: Separating Traits, Episodes, and Confounders Over Time

A longitudinal formulation organizes a patient's entire psychiatric history into layers: what has been present since childhood (developmental), what emerged after specific events (acquired), what comes and goes in discrete episodes (episodic), and what varies with substances, medications, or life circumstances (context-dependent). This is more useful than a cross-sectional diagnosis because it explains not just what the patient has now, but why they have it and how different components interact. A patient with ADHD since childhood, borderline personality traits since adolescence, recurrent depressive episodes since their 20s, and substance-induced mood instability currently has four distinct layers — each requiring different treatment.

What was present before age 12, before trauma, before substance use? ADHD, autism spectrum features, intellectual disability, learning disorders, and early temperamental extremes belong in this layer. These are trait-like and lifelong. Key evidence: school records showing inattention, hyperactivity, or social difficulties; family reports of early childhood behavior; academic underperformance despite adequate intelligence. If ADHD symptoms were clearly present before trauma onset, ADHD is a primary condition rather than a trauma consequence — even if trauma later worsened concentration.

Personality develops from adolescence through early adulthood. Personality traits — relational patterns, identity coherence, emotional regulation style, and interpersonal functioning — are relatively stable over years. Borderline personality features (abandonment fear, identity disturbance, idealization-devaluation, self-harm as regulation), narcissistic features (grandiosity, entitlement, empathy deficits), avoidant features (social inhibition, feelings of inadequacy), and other personality patterns belong in this layer. These are not episodes — they are the patient's characteristic way of being in the world. They may fluctuate in intensity with stress but do not remit into periods of normal functioning the way mood episodes do.

Mood episodes (depression, hypomania, mania), psychotic episodes, anxiety disorder onset, and PTSD onset are episodic — they have a beginning, middle, and end. Key questions for this layer: Can you identify when this episode started? Was there a period of relative normalcy before it? Is it qualitatively different from the patient's baseline? Does it resolve (with or without treatment) and return to baseline? A depressive episode that started 3 months ago in someone who was previously functional is different from lifelong dysthymia in someone who has never felt good. Bipolar episodes should be documented with approximate dates, duration, triggers (if any), and severity.

Some symptoms are directly attributable to current circumstances rather than a primary disorder. Active substance use, medication effects, sleep deprivation, acute life stressors, medical illness, and social isolation all produce psychiatric symptoms that may resolve when the context changes. This layer is important because treating context-dependent symptoms as a primary disorder leads to overdiagnosis and overmedication. A patient who is depressed, anxious, and unable to concentrate during a divorce, while drinking heavily and sleeping 4 hours per night, may not have depression, anxiety, or ADHD — they may have a terrible life situation. Address the context before diagnosing.

For a complex patient, the formulation should explicitly state what belongs in each layer. Example structure: Developmental: probable ADHD (present before age 12, before trauma, consistent with family history). Personality: borderline features (abandonment fear, identity instability, self-harm since adolescence — consistent with early neglect and chaotic caregiving). Episodic: possible bipolar II (discrete periods of elevated energy and decreased sleep need, some appearing autonomous — but confounded by substance use and medication changes). Context-dependent: current cannabis use contributes to amotivation and paranoia; sleep deprivation worsens all layers. This structure makes it clear which symptoms are targets for medication (episodic mood), which are targets for psychotherapy (personality), which require behavioral intervention (substances, sleep), and which were always present and may need accommodation (ADHD).

Collapsing everything into one diagnosis: a complex patient gets one label (usually bipolar or BPD) and everything is explained through that lens. This leads to treatments that address one layer while ignoring others. Ignoring developmental history: diagnosing ADHD as bipolar because of current impulsivity, without asking whether impulsivity was present in childhood before any mood symptoms. Treating personality as episodic: putting a BPD patient on mood stabilizer after mood stabilizer because each emotional crisis is interpreted as a mood episode rather than a personality-driven reaction. Ignoring substances: diagnosing bipolar disorder during active stimulant use without establishing whether mood episodes occur during sobriety. Anchoring on family history: family history of bipolar I raises the prior probability but does not determine the diagnosis. A first-degree relative with bipolar I increases risk approximately 10-fold, but the base rate is 1% — meaning 90% of first-degree relatives do NOT develop bipolar disorder.