The Lithium Tremor That Means Toxicity, Not a Side Effect

The PMHNP boards love the tremor question because most students treat “tremor” as one thing. It isn’t. The trap is handing you a patient on lithium with a tremor and seeing whether you reassure them or panic appropriately. A new coarse, slow tremor on lithium is a toxicity flag — not a benign side effect — and it changes the whole management. Sort every tremor by when it appears and the question answers itself.

Sort by timing: rest vs. action

The single discriminator is whether the tremor shows up at rest or with movement.

  • Resting tremor — present at rest, quiets with action. This is drug-induced parkinsonism, classically from dopamine blockers (antipsychotics). The move is a medication review of the antipsychotic.
  • Postural/action tremor — appears when the limb is held against gravity or moving. This is the lithium bucket, plus valproate, essential tremor, and physiologic/anxiety tremor.

Get that split first. It decides everything downstream.

The benign lithium tremor — but verify

A fine postural/action tremor on lithium is common and often benign. That’s true. But “often benign” is not “ignore it.” Before you reassure the patient, check a lithium level — because the next paragraph is the reason.

The red flag

Here’s the finding the boards are testing: a NEW COARSE, SLOW action tremor on lithium = lithium toxicity until proven otherwise.

Fine and longstanding is one thing. New, coarse, and slow is a different animal. When you see it:

  • Hold the dose
  • Draw a STAT lithium level
  • Assess the toxicity cluster: GI upset and vomiting, ataxia, dysarthria, confusion, hyperreflexia — progressing to seizures and coma at the severe end

The coarse tremor travels with that neuro-GI cluster. That’s what separates it from the benign fine tremor that travels alone.

What the numbers mean — and don’t

Therapeutic lithium is roughly 0.6–1.2 mEq/L. A level >1.5 signals toxicity; >2.0 is generally moderate-to-severe concern.

But don’t anchor to the number alone. Clinical severity depends on symptoms, chronicity, renal function, age, and neurologic findings. A chronically toxic older adult with poor renal clearance can be very sick at a level that looks only modestly elevated. Treat the patient and the picture, not just the lab value.

Treating the action tremor (when it’s NOT toxicity)

For an action/postural tremor that isn’t a toxicity signal — lithium fine tremor, valproate tremor, essential tremor, physiologic/anxiety tremor — the answer is propranolol and/or dose reduction.

The mirror-logic traps

This is where the exam scores points off reflexes. Keep these straight:

  • Propranolol is for action tremor, NOT parkinsonian rest tremor. Don’t beta-block a resting tremor and expect a fix.
  • Anticholinergics are for parkinsonism and dystonia, NOT akathisia. Different problem, different drug.

And one more clean-up: AIMS is a dyskinesia screening tool — it’s not the trigger here. The trigger for the lithium tremor is med and level review, not a dyskinesia scale.

Why this is easy to miss

“Tremor on lithium” pattern-matches to “expected side effect, reassure,” and most of the time that’s even right. The boards build the question around the exception — the new, coarse, slow tremor riding with ataxia and dysarthria — to see if you’ll reflexively reassure a toxic patient. Read the quality and the company it keeps, not just the word “tremor.”

On the exam: new coarse, slow action tremor on lithium → hold the dose, STAT level, look for the ataxia/dysarthria/confusion cluster — that’s toxicity, not a side effect.


Meridian’s board-prep course drills the timing-and-management discriminations the exam hides its answers inside — with worked cases and MSE video demos. Explore the board-prep course.


← All articles