The 38-year-old on the stroke service generates two reflexes, and both are wrong. The first is disbelief — re-reading the MRI, hunting for a mimic, because "he's too young for this." The second is the shotgun — sending every rheumatologic, hematologic, and genetic test the laboratory offers on hospital day one. The first reflex delays treatment. The second generates a pile of weakly positive results that will haunt the patient for years.

What a young stroke actually demands is a tiered search: everyone gets the core; the story, the vessels, and the infarct pattern decide who earns the next tier.1

The etiology spectrum: what is actually common under 50

The distribution is the argument for the workup. In young-adult cohorts, cervical artery dissection is the leading identified arterial mechanism — up to a quarter of ischemic strokes under 50 — followed by cardioembolism (including PFO-associated stroke), early-onset atherosclerosis and small vessel disease climbing steeply through the 40s, substance-associated stroke (cocaine, amphetamines), and a long tail of rarities.12 Even here, the CADISP data are humbling about "healthy young people": hypertension was over-represented in dissection patients versus referents.3 And a meaningful fraction remain cryptogenic after a full evaluation — a label that is a prompt to escalate, not a diagnosis to accept on day three.

The dissection story — neck pain, partial Horner, a TIA that "went away" — has its own chapter: cervical artery dissection, including the CADISS/TREAT-CAD antithrombotic decision. This page is the map around it.

The tiered workup

Tier Who What
1 — EveryoneEvery young ischemic strokeArch-to-vertex CTA/MRA including the neck (fat-sat T1 if dissection suspicion persists), MRI, echocardiogram with agitated saline, rhythm monitoring, lipids, A1c, tox screen, pregnancy test where relevant
2 — Pattern-drivenCryptogenic after tier 1; multi-territory infarcts; venous-pattern events; systemic featuresProlonged rhythm monitoring, TEE where it changes management, antiphospholipid antibodies, inflammatory markers, hypercoagulable panel timed and targeted, CSF/vessel-wall imaging when vasculitis is genuinely on the table
3 — Door-specificThe zebra whose door the patient walked throughNOTCH3 testing (CADASIL pattern), catheter angiography (moyamoya), head-to-pelvis CTA (FMD), genetics/metabolic testing (Fabry, mitochondrial) with counseling

The full checklist version — with the traps at each step — lives on the stroke workup page; the mechanism you land on gets classified on the TOAST page, where most of this cluster files under "other determined."

The PFO problem: culprit or bystander

A patent foramen ovale is present in roughly one in four adults, which means most PFOs found after stroke are innocent. The RoPE index formalizes the intuition: the younger the patient and the fewer the vascular risk factors, the more likely the PFO is pathogenic — a superficial infarct pattern adds to it.4 Score it on our RoPE calculator before the closure conversation, because the score frames whose PFO is worth closing: guidelines now support closure in selected patients under 60 with a nonlacunar, otherwise-cryptogenic infarct and a high-probability PFO after genuine shared decision-making.5 The order of operations matters — a PFO does not end the workup that would have found the dissection or the atrial fibrillation.

Finding a PFO is the beginning of an argument, not the end of a workup.

Thrombophilia and APS: test the right blood at the right time

The inherited thrombophilias (factor V Leiden, prothrombin mutation) are principally venous risk factors; their arterial payoff is low outside paradoxical embolism, which is why the panel belongs to tier two and to patients whose story earns it. The exception with real arterial teeth is antiphospholipid syndrome — young stroke plus livedo, pregnancy morbidity, thrombocytopenia, or prior thromboses should trigger the antibody panel, with the persistence-at-12-weeks rule and the warfarin-not-DOAC treatment wrinkle we walk through on the APS criteria page. Acute-phase and anticoagulated blood distorts several of these assays; drawing the full panel on hospital day one is how patients acquire diagnoses they do not have.

The zebras, by their doors

  • CADASIL — recurrent lacunar strokes without hypertension, migraine with aura, mood change, family history, and the anterior-temporal-pole white matter signature: the door to NOTCH3 testing, covered in the CADASIL guide.
  • Moyamoya — recurrent anterior-circulation events with bilateral distal ICA narrowing and hazy collaterals: the door to catheter angiography and a revascularization conversation, covered in the moyamoya guide.
  • Fibromuscular dysplasia — dissection (especially multivessel or recurrent), renovascular hypertension, pulsatile tinnitus, string-of-beads imaging: the door to head-to-pelvis screening, covered in the FMD guide.
  • Vasculitis and endocarditis — multi-territory infarcts with systemic inflammation or fever: the door to CSF, vessel-wall imaging, blood cultures, and the Duke criteria.
  • Pregnancy and the puerperium — a young-stroke context of its own (preeclampsia spectrum, RCVS, venous thrombosis) where timing is the diagnostic clue.

The bottom line

Treat the young stroke like a stroke first — reperfusion decisions do not wait for etiology. Then search in tiers: neck vessels and heart for everyone, because dissection and embolism lead the list; escalate to the timed, targeted second tier when the first is clean; and open the zebra doors only when the pattern knocks. A young patient's diagnosis sets fifty years of prevention — the workup is worth doing precisely once, and well.