Clinical Tool

Differential Diagnosis Decision Tree

Anxiety Differential Diagnosis Decision Tree

Quick Reference Tool | PsychHQ Source: Module 2, Evaluation & Diagnosis (Lessons 4, 5, 6) | Last Updated: March 2026


Core Principle

The DSM-5-TR requires that a primary anxiety diagnosis is not attributable to the physiological effects of a substance, medication, or another medical condition. Yet diagnostic overshadowing — where somatic symptoms of medical illness are prematurely attributed to a psychiatric disorder — remains a pervasive clinical failure. Your job before diagnosing primary anxiety: rule out medical mimics, screen for substance-induced causes, and differentiate from overlapping psychiatric conditions.


Step 1: Medical Rule-Outs

Standard Workup for New-Onset Anxiety

Based on AAFP, APA, and AACAP recommendations:

Component Purpose
CBC (complete blood count) Rule out anemia, systemic infection
CMP (comprehensive metabolic panel) Glucose, electrolytes, renal/hepatic function
TSH + free T4 Rule out hyper/hypothyroidism
Urinalysis General screening
Urine drug screen Identify covert substance use or withdrawal
Medication and substance review Rule out iatrogenic causes (caffeine, stimulants, corticosteroids, decongestants)

Advanced testing (extended cardiac monitoring, brain MRI, EEG, fractionated metanephrines, infectious titers) is reserved for patients with specific clinical red flags — not applied universally.

Medical Conditions That Mimic Anxiety

Condition How It Mimics Anxiety When to Suspect What to Order
Hyperthyroidism Tachycardia, diaphoresis (sweating), tremor, hypervigilance — virtually mirrors panic/GAD Weight changes, cold/heat intolerance, goiter, menstrual changes TSH + free T4 (already in standard workup)
Hashimoto's thyroiditis (even when euthyroid — normal thyroid hormone levels) OR (odds ratio — how much more likely the condition is) 2.32 for anxiety (Siegmann et al., 2018, JAMA Psychiatry); euthyroid Hashimoto's: OR 2.52 (Romero-Gomez et al., 2024). Likely mechanism: thyroid autoantibodies cause neuroinflammation independent of hormone levels Treatment-resistant anxiety with no other etiology; family history of autoimmune disease TPO and TG antibodies (thyroid peroxidase and thyroglobulin — markers of autoimmune thyroid activity; not yet in standard guidelines, but emerging evidence supports testing in treatment-resistant cases)
Pheochromocytoma Massive catecholamine (adrenaline and related stress hormone) surges produce episodes indistinguishable from severe panic. Rare (0.8–3 per million), but up to 50% are unrecognized at autopsy Paroxysmal or treatment-resistant hypertension; panic with facial pallor (not flushing); new-onset panic in older adult without psychiatric history; paradoxical BP responses to beta-blockers or TCAs Plasma free or urinary fractionated metanephrines (blood or urine test measuring breakdown products of adrenaline)
SVT (supraventricular tachycardia) 67% of patients diagnosed with PSVT (paroxysmal SVT) met full panic disorder criteria (Lessmeier et al., 1997). 55% unrecognized after initial evaluation; median diagnostic delay 3.3 years. After catheter ablation, only 4% still met panic criteria Sudden-onset palpitations; episodes self-terminating; women 2x more likely to have cardiac symptoms attributed to psych (65% of women vs 32% of men) Standard ECG + prolonged ambulatory monitoring (14–30 day event monitors — 24-hour Holter is often insufficient for sporadic arrhythmias)
Asthma / COPD Dyspnea triggers the brain's suffocation alarm; hyperventilation worsens airway resistance. Anxiety in 16–52% of asthma, up to 40% of COPD patients. COPD + anxiety doubles exacerbation risk (HR — hazard ratio — 2.10) Known respiratory disease with worsening anxiety; dyspnea disproportionate to pulmonary function Pulmonary function testing; distinguish anxiety-driven hyperventilation from organic dyspnea
Vestibular disorders RR (relative risk) 1.51 for anxiety. Vestibular nuclei connect directly to the brain's fear circuitry Dizziness, vertigo, or imbalance with co-occurring anxiety; anxiety that worsens with head movement or positional changes Vestibular function testing
Temporal lobe epilepsy Ictal fear (fear occurring during a seizure episode) closely mimics panic attacks. Psychiatric comorbidity reaches 79% in pediatric TLE; anxiety in 14–23% Very brief episodes (<1–2 min); associated staring or lip-smacking; post-episode confusion; failure to respond to standard anxiety treatment EEG

Mitral valve prolapse — a historical link now debunked. Early studies reported elevated MVP prevalence (up to 35%) in panic patients. Modern controlled echocardiographic studies with blinded evaluation show no significant difference in MVP prevalence between panic disorder, social anxiety, and healthy controls. The 2020 ACC/AHA Valvular Heart Disease Guidelines make no mention of anxiety as an MVP feature. If a patient with known MVP has panic attacks, treat the panic on its merits.

Red Flags That Expand the Workup

Any of these should halt a psychiatric diagnosis and trigger further investigation: new-onset anxiety after age 35 with no psychiatric history or psychosocial stressor; somatic symptoms disproportionate to the psychological picture; panic with facial pallor or treatment-resistant hypertension; episodes under 2 minutes with post-episode confusion or automatisms (involuntary movements like lip-smacking); overnight onset of OCD with behavioral regression in a child (consider PANS/PANDAS — see Module 2, Lesson 4); complete failure of standard anxiety treatment over 12+ weeks.


Step 2: Substance and Medication Screen

The DSM-5-TR recognizes Substance/Medication-Induced Anxiety Disorder (SMIAD) as a distinct category. Identifying the offending agent often allows complete symptom resolution without lifelong psychotropic management.

Agent Evidence Clinical Action
Caffeine Overall SMD (standardized mean difference — a measure of effect size where >0.8 is large) 0.94 for anxiety (Liu et al., 2024, 8 RCTs). Sharply non-linear dose-response: <400 mg/day SMD 0.61; ≥400 mg/day SMD 2.86 — nearly fivefold increase Quantify intake in milligrams, not "cups." Energy drinks and pre-workout supplements easily exceed 400 mg per serving. If intake exceeds 400 mg and patient has new-onset panic/palpitations, reduce caffeine and observe 2–3 weeks before diagnosing primary anxiety
Cannabis (THC) Both 10 mg and 25 mg THC doses increased anxiety in double-blind crossover studies (Spindle et al., 2018). Women report greater anxiety increases than men (Sholler et al., 2021). Regular users show blunted effects due to tolerance; naive users at highest risk Ask specifically about cannabis use. THC is anxiogenic (anxiety-producing). CBD may have anxiolytic properties but optimal dosing is completely unestablished (study doses ranged 14.3–1,000 mg/day)
Therapeutic stimulants RR (relative risk) 1.34 for overall adverse events vs placebo (2025 meta-analysis, 93 RCTs, 11,034 patients), including anxiety among most common For ADHD + comorbid anxiety: consider non-stimulants first (atomoxetine, viloxazine, guanfacine). If stimulants necessary, extended-release methylphenidate with cautious titration
Pseudoephedrine Indirect sympathomimetic: tachycardia, tremors, jitteriness identical to panic. Patients rarely report OTC medications during intake Ask explicitly about cold, sinus, and allergy medication use. Contraindicated with MAOIs (fatal hypertensive crisis risk)
Corticosteroids Mood disturbances ~17%, anxiety 8%, psychotic features 5%. >80 mg/day prednisone equivalent: 18.6% severe psychiatric symptoms vs 1.3% on low doses. Acute onset within 3–4 days; chronic exposure (>90 days) produces gradual anxiety and cognitive decline Cautious dose reduction when possible; morning-only dosing of short-acting preparations to mimic diurnal cortisol rhythm
Alcohol withdrawal Mild anxiety/tremor 6–12 hours after cessation; peak danger 24–72 hours (seizures, delirium tremens); protracted baseline anxiety may persist months Screen with AUDIT (Alcohol Use Disorders Identification Test). Timeline of cessation relative to anxiety onset is diagnostic
Benzodiazepine discontinuation Rebound anxiety transiently exceeds pre-treatment baseline. Worse with short-acting agents (alprazolam > diazepam). Distinctive features: tinnitus, internal vibrations, "only fear and anxiety" Distinguish from relapse: discontinuation starts within days and includes neurological symptoms (brain zaps, paresthesias) not part of original disorder. Relapse develops gradually over weeks
SSRI/SNRI discontinuation FINISH mnemonic: flu-like, insomnia, nausea, imbalance, sensory disturbances, hyperarousal. Onset within 2–4 days; resolves within 1–2 weeks. Paroxetine and venlafaxine highest risk; fluoxetine lowest Rapid onset + neurological symptoms = discontinuation. Gradual onset + original symptom pattern = relapse

The temporal relationship is the cornerstone. Primary anxiety is diagnosed when symptoms clearly precede any substance use, or persist for 1+ month after verified abstinence. SMIAD is diagnosed when anxiety developed during or within one month of intoxication or withdrawal, and the substance is pharmacologically capable of producing the observed symptoms.


Step 3: Psychiatric Differential Diagnosis

Condition How It Overlaps With Anxiety Key Differentiating Feature Practical Discriminator
ADHD Both produce inattention, restlessness, executive dysfunction. 34–50% of adults with ADHD have comorbid anxiety (Frontiers in Psychiatry, 2025). ASRS hyperactivity items ("difficulty relaxing," "driven by a motor") load onto anxiety factors in factor analysis (Spann et al., 2024) Inattention mechanism: ADHD = mind pulled outward to more stimulating things (often unaware). Anxiety = mind locked onto worry (aware and distressed). Impulsivity is the strongest discriminator: in ADHD it's pervasive and mood-independent; in anxiety it's affect-dependent, occurring only when overwhelmed "Are you distracted by a racing mind full of worries, or does your mind drift when a task is boring?" "Do you act impulsively even when calm, or only when overwhelmed?"
ASD ~50% of ASD patients have social anxiety. High-functioning adolescents with ASD can be misread as having standard SAD Motivation for avoidance: SAD = fear of negative evaluation (intact social skills, blocked by hyperarousal). ASD = cognitive exhaustion from manually decoding social rules + intolerance of uncertainty + sensory processing differences. Eye-tracking: social anxiety = faster orienting away from eyes (fear-based); ASD = delayed orienting to eyes (reduced social attention) (Kleberg et al., 2017) "Is social interaction frightening because of judgment, or exhausting because of the effort of reading people?"
Trauma / PTSD Avoidance, hyperarousal, and fear appear in both. Many patients with trauma meet GAD criteria Fear structure: Anxiety = generalized, anticipatory, focused on uncertain future threats; expands over time through overgeneralization. PTSD = anchored to a specific event; avoidance maps to trauma-related stimuli. Re-experiencing (flashbacks, intrusive memories, nightmares about the event) distinguishes PTSD from anxiety, which does not feature re-experiencing "Can you identify when this started? Was there a specific event?" Probe for somatic avoidance: unexplained shortness of breath, chronic pain, or body heat as autonomic trauma reactions
Depression Co-occurs in >50% of anxious outpatients. PHQ-9 (Patient Health Questionnaire-9, a depression screener) alone misses 34.3% of comorbid anxiety (PHQ-ADS — combined anxiety-depression scale — validation) When both are present and severe depression includes active suicidality: treat depression aggressively as primary target. For milder dual presentations, treating primary anxiety with CBT often improves depressive symptoms secondarily Always pair GAD-7 with PHQ-9. Don't rely on either alone
Normal developmental anxiety Children show age-appropriate fears at every stage: separation anxiety in infancy, dark/creatures in toddlerhood, social evaluation in adolescence Three criteria cross the line to pathological: (1) reaction is disproportionate and decoupled from realistic danger, (2) drives persistent behavioral avoidance of normative experiences, (3) causes functional impairment beyond what is developmentally expected Distress alone is insufficient. Evaluate persistence, unreasonableness, and actual functional impact over time — not a cross-sectional snapshot
GAD vs. illness anxiety vs. somatic symptom disorder All feature health-related worry and somatic complaints GAD: health worry is one dimension of diffuse worry that also spans finances, relationships, and work. Somatic symptom disorder: distressing physical symptoms dominate, with disproportionate thoughts/behaviors about those symptoms for 6+ months. Illness anxiety: obsessive worry about having a serious illness despite normal findings, with actual symptoms absent or only mild Rule out primary anxiety, panic, and depression before confirming SSD or IAD

Step 4: Diagnostic Biases to Check

Gender: Men externalize anxiety as aggression, irritability, risk-taking, and substance use. Standard instruments fail to capture these male-specific patterns. This attribution bias contributes to men's 4x higher suicide completion rate (Call & Shafer, 2018).

Culture: Ataques de nervios (Latinx populations — shouting, trembling, dissociative features; 15% of adults in a Puerto Rican study reported them, 63% met criteria for underlying psychiatric disorder; frequently misdiagnosed as psychosis). "Thinking too much" (Sub-Saharan Africa — ruminative intrusive thoughts with physical decline). Taijin kyofusho (Japan — fear that one's presence will offend others, not fear of personal embarrassment).

Age: Anxiety in older adults (prevalence 1.2–15%) is routinely misattributed to cardiovascular disease, respiratory illness, or "normal aging." The BAI performs poorly in elderly (sensitivity 0.70, specificity 0.60 due to somatic weighting). Use the GAI-20 instead (sensitivity 0.89, specificity 0.80).

Over-diagnosis risk in children: Anxious parents over-report and catastrophize their child's symptoms. Multi-informant assessment (parent + teacher + child) is the primary safeguard. If anxious parenting and accommodation are the main drivers, the intervention target is parental behavior, not the child's pharmacotherapy.

Structured beats unstructured. In every study, structured instruments outperform unassisted clinical judgment. Use them.


Differential Diagnosis Summary Table

Condition Mimicry Mechanism Key Distinguishing Feature What to Order/Ask
Thyroid dysfunction Altered metabolism mimics panic/GAD Physical signs; reversible with treatment TSH + free T4; TPO/TG antibodies if treatment-resistant
Pheochromocytoma Catecholamine surges mimic panic Facial pallor, treatment-resistant hypertension Fractionated metanephrines
SVT Cardiac arrhythmia produces panic-like episodes Self-terminating; post-ablation resolution Prolonged ambulatory monitoring
Caffeine Sympathomimetic activation Dose-dependent; resolves with reduction Quantify mg intake
Cannabis (THC) Anxiogenic at all studied doses Correlates with use; naive users at highest risk Direct substance history
ADHD Shared inattention and restlessness Mood-independent impulsivity; lifelong course "Distracted by worry or by boredom?"
ASD Shared social difficulty Social exhaustion vs fear of judgment; sensory differences "Frightening or exhausting?"
PTSD Shared avoidance and hyperarousal Anchored to specific event; re-experiencing symptoms Trauma history; "When did this start?"
Depression >50% co-occurrence Joint screening mandatory; treat severe depression first GAD-7 + PHQ-9 together
Normal development Age-appropriate fears cause distress Disproportionate + avoidance + functional impairment = pathological Evaluate persistence over time

Key References: Vermani et al. (2011, GP detection rates); Siegmann et al. (2018, Hashimoto's-anxiety, JAMA Psychiatry); Romero-Gomez et al. (2024, euthyroid Hashimoto's); Lessmeier et al. (1997, SVT-panic overlap); Liu et al. (2024, caffeine meta-analysis); Spindle et al. (2018, THC-anxiety); Spann et al. (2024, ASRS factor analysis); Kleberg et al. (2017, eye-tracking SAD vs ASD); Call & Shafer (2018, male anxiety expression); Guarnaccia et al. (1993, ataques de nervios); McKnight et al. (2016, symptom-function correlation); Atchison et al. (2024, GAI-20)