Lithium: The Good, the Bad, and the Ugly

For educational purposes only. This post is not medical advice and is not intended to guide treatment decisions. Please speak with your medical provider about your own care.

Lithium is the oldest drug still sitting at the top of a modern psychiatric treatment algorithm. It was in use before the FDA required proof of efficacy, it costs almost nothing, no company has any reason to market it, and it remains first-line maintenance therapy for bipolar disorder in essentially every international guideline. [1][2][3]

It is also the drug psychiatry argues about most. Prescribing has been falling for decades, patients who need it often wait years to be offered it, and the reasons are not mysterious — lithium asks a great deal of both the person taking it and the person prescribing it.

Here is the honest ledger.

The Good

It works, and it works on both poles. Network meta-analysis puts the relapse risk ratio at 0.62 versus placebo, with protection against manic and depressive recurrence. [2][3] That combination is unusual. Most agents that reliably prevent mania are weaker on the depressive side, which matters because depression accounts for the majority of symptomatic time in bipolar disorder.

It may change the disease, not just the symptoms. Bipolar disorder increasingly looks like a progressive condition, with each episode making the next one more likely and treatment response harder to achieve. Lithium acts on a striking number of the mechanisms and surrogate markers implicated in that progression, and it has documented neuroprotective effects — which is why it is now discussed in the language of disease-modifying therapy rather than symptom control. [4] Register data also suggest that starting it earlier in the illness course produces better response rates than starting it after episodes have accumulated. [5]

Long-term outcomes favor it. Across decades of observational and trial data, patients maintained on lithium-based regimens do better than those on non-lithium regimens on hospitalization, relapse, and overall course. [4][1]

The suicide question — genuinely good, genuinely contested. For years the standard claim was that lithium reduces suicide, anchored to a 2013 meta-analysis reporting an odds ratio of 0.13 versus placebo. [6] That consensus has since been challenged. A large VA randomized trial designed specifically to test lithium for preventing repeat suicide-related events was stopped early for futility, finding no preventive effect. [7] A subsequent meta-analysis restricted to modern trials concluded the evidence is inconclusive. [8] The fair summary today: there is a plausible, long-observed signal, and it is weaker and less settled than it was widely taught to be.

It is cheap and it is knowable. Lithium costs pennies. Its blood level is measurable, its therapeutic window is defined, and after seventy-five years of use there are very few surprises left in it. That is not true of most psychotropics.

The Bad

These are the day-to-day burdens — rarely dangerous, frequently the reason people stop.

Tolerability. Fine hand tremor, increased thirst and urination, weight gain, gastrointestinal upset, cognitive dulling or "flatness," acne, and psoriasis flares are all common. [9] The cognitive complaint deserves particular attention because it is often dismissed, and because it is one of the most common reasons patients quietly stop taking a drug that was working.

Thyroid. Hypothyroidism is substantially more common on lithium — reported at 13.9% versus 1.7%, an odds ratio near 5.8. [1][3] It is manageable with levothyroxine and is not usually a reason to stop lithium, but it requires that someone is actually checking.

Parathyroid and calcium. Hyperparathyroidism and hypercalcemia occur more often on long-term lithium, which is why calcium belongs on the monitoring panel alongside thyroid and kidney function. [1][3]

The monitoring commitment. Baseline and ongoing renal, thyroid, and calcium testing, plus periodic serum lithium levels, with maintenance targets typically around 0.6–0.8 mEq/L. [1][9] This is not an onerous protocol on paper. In practice it requires a stable care relationship, reliable lab access, and a patient who can get to phlebotomy every few months for years — and it is a real reason lithium gets passed over for drugs that require nothing.

Interactions that catch people off guard. Lithium is cleared by the kidney and handled like sodium. NSAIDs, ACE inhibitors, ARBs, and thiazide diuretics all raise levels. So does anything that causes volume depletion: a stomach bug, a heat wave, a hard workout, a new low-sodium diet. Many lithium toxicity cases begin with something entirely ordinary.

It does not suit everyone. Lithium is meaningfully less effective in dysphoric or mixed mania, rapid-cycling course, prominent psychotic features, and comorbid alcohol use. [9] The classic best responders are people with an episodic course, full recovery between episodes, and a family history of lithium response.

The Ugly

These are the ones that justify the reputation.

Toxicity is a narrow-window problem. The gap between a therapeutic level and a toxic one is small. Levels above roughly 1.5 mEq/L bring coarse tremor, vomiting, ataxia, and confusion; above about 2.5 mEq/L the picture can progress to seizures, arrhythmia, renal failure, and coma, sometimes requiring dialysis. Chronic toxicity can develop at levels that look acceptable on paper, particularly in older adults or during intercurrent illness. [9]

SILENT. A subset of patients who survive lithium toxicity are left with permanent neurological damage — the syndrome of irreversible lithium-effectuated neurotoxicity, most often persistent cerebellar dysfunction with ataxia, dysarthria, and nystagmus that does not resolve when the drug is cleared. [9] It is uncommon. It is also the reason a lithium level is not a formality.

The kidney, slowly. Long-term lithium is associated with an increased incidence of chronic kidney disease, with risk rising with treatment duration, higher levels, older age, prior episodes of toxicity, and lower baseline function. Progression to end-stage kidney disease appears to be rare, and in more recent cohorts with better monitoring, rarer still — but it happens. [10][11] The honest framing is a gradual, monitorable decline in most people and a small tail of serious outcomes, not an inevitability.

Pregnancy. First-trimester lithium exposure increases cardiac malformation risk, and the effect is dose-dependent: adjusted risk ratios of 1.11 at 600 mg/day or less, 1.60 at 601–900 mg, and 3.22 above 900 mg. [12] In absolute terms this is roughly two cases per 100 live births versus one in unexposed pregnancies — substantially smaller than the alarming figures from 1970s registry data, but not nothing. [12] The corresponding risk of leaving bipolar disorder untreated through pregnancy and the postpartum period is also not nothing, which makes this one of the harder conversations in the field.

Stopping badly is worse than never starting. This is the part patients are least often warned about. After long-term maintenance, abrupt discontinuation is followed by a median of about 4 months to recurrence, versus about 20 months when lithium is tapered gradually over several weeks — a fivefold difference. In the first year, monthly recurrence rates run 6.5% after rapid discontinuation versus 2.3% after gradual. [13] There are also reports that a subset of patients who stop abruptly do not respond as well when lithium is restarted, though this remains debated. [13] The practical implication is blunt: if lithium is going to be stopped, it should be stopped slowly and deliberately, not because a prescription lapsed or a refill fell through.

The bottom line

Lithium is not a gentle drug and it is not a safe default. It demands lab monitoring, a reliable prescriber relationship, and a patient willing to stay engaged for years. Those demands are the reason it is underused, and they are legitimate.

But the reasons it has survived seventy-five years are also legitimate: broad efficacy across both poles of the illness, evidence for altering the disease course rather than suppressing symptoms, a long-term outcome record no newer agent has matched, and a risk profile that — unlike most of psychiatry's newer options — is thoroughly mapped.

The right question is almost never "is lithium safe?" It is "is this the right patient, is this the right moment in the illness, and is the monitoring infrastructure actually in place?" When the answer to all three is yes, lithium remains the strongest tool available. When it isn't, none of the above makes it a good idea.

Disclaimer: This article is provided for educational and informational purposes only. It is not intended as medical advice, diagnosis, or treatment, and should not be used as a substitute for consultation with a qualified healthcare professional. Lithium requires individualized dosing, laboratory monitoring, and careful attention to drug interactions and hydration status. Do not start, stop, or change the dose of any medication without speaking with your medical provider. Abrupt discontinuation of lithium carries specific risks described above — if you are considering stopping, talk to your prescriber about tapering safely.

References

1. Nierenberg AA, Agustini B, Köhler-Forsberg O, et al. Diagnosis and treatment of bipolar disorder: a review. JAMA. 2023.

2. Carvalho AF, Firth J, Vieta E. Bipolar disorder. The New England Journal of Medicine. 2020.

3. Malhi GS, Gessler D, Outhred T. The use of lithium for the treatment of bipolar disorder: recommendations from clinical practice guidelines. Journal of Affective Disorders. 2017.

4. Post RM, Li VW, Berk M, Yatham LN. Lithium as a disease-modifying drug for bipolar disorder. The Lancet Psychiatry. 2025.

5. Kessing LV, Vradi E, Andersen PK. Starting lithium prophylaxis early v. late in bipolar disorder. The British Journal of Psychiatry. 2014.

6. Cipriani A, Hawton K, Stockton S, Geddes JR. Lithium in the prevention of suicide in mood disorders: updated systematic review and meta-analysis. BMJ. 2013;346:f3646.

7. Katz IR, Rogers MP, Lew R, et al. Lithium treatment in the prevention of repeat suicide-related outcomes in veterans with major depression or bipolar disorder: a randomized clinical trial. JAMA Psychiatry. 2022;79:24–32.

8. Nabi Z, Stansfeld J, Plöderl M, Wood L, Moncrieff J. Effects of lithium on suicide and suicidal behaviour: a systematic review and meta-analysis of randomised trials. Epidemiology and Psychiatric Sciences. 2022;31:e65.

9. Baldessarini RJ, Tondo L, Vázquez GH. Pharmacological treatment of adult bipolar disorder. Molecular Psychiatry. 2019.

10. Risk of chronic kidney disease in individuals on lithium therapy in Iceland: a nationwide retrospective cohort study. The Lancet Psychiatry. 2024.

11. Boivin E, Le Daré B, Bellay R, Vigneau C, Mercerolle M, Bacle A. Long-term lithium therapy and risk of chronic kidney disease, hyperparathyroidism and hypercalcemia: a cohort study. International Journal of Bipolar Disorders. 2023.

12. Patorno E, Huybrechts KF, Bateman BT, et al. Lithium use in pregnancy and the risk of cardiac malformations. The New England Journal of Medicine. 2017;376:2245–2254.

13. Baldessarini RJ, Tondo L, Faedda GL, Suppes TR, Floris G, Rudas N. Effects of the rate of discontinuing lithium maintenance treatment in bipolar disorders. The Journal of Clinical Psychiatry. 1996;57:441–448.