Lithium salts

Article written and reviewed by Dr Luca Moretti

Lithium is a mood stabilizer with a narrow therapeutic index and a well established role, particularly in bipolar disorder. In psychiatry it is used in mania and, especially, for prevention of recurrence in bipolar disorder; it can also be added to an antidepressant in major depressive disorder when response is insufficient. These uses are not equivalent in strength of evidence, target serum lithium levels or duration of treatment and should therefore be distinguished. The Italian Summary of Product Characteristics for lithium carbonate retains historically authorized wording, while contemporary clinical practice also defines indications, plasma levels and monitoring through updated international guidelines.

The feature that most strongly determines clinical use is the close relationship between systemic exposure, efficacy and toxicity. Lithium is not metabolized by the liver and is eliminated almost entirely by the kidneys; changes in glomerular filtration, sodium intake, circulating volume, or the introduction of drugs that reduce its clearance can rapidly increase the plasma concentration. For this reason, appropriate prescribing involves not only selecting the dose but also standardized serum lithium measurement, patient education, interaction management and surveillance of renal and thyroid function and calcium metabolism.

Pharmacology, mechanisms of action and pharmacokinetics

Lithium is a monovalent ion and does not have a single pharmacological target that by itself explains its mood stabilizing effect. Experimental models indicate modulation of multiple intracellular signaling systems, including the phosphoinositide cycle and the enzyme glycogen synthase kinase 3 (GSK 3), together with effects on second messengers, gene expression, monoaminergic and glutamatergic transmission, circadian rhythms, mitochondrial function and pathways involved in synaptic plasticity. These mechanisms are biologically plausible and documented at several experimental levels, but they do not allow clinical efficacy to be attributed to a single molecular cascade or a mechanistic biomarker to be used to predict an individual patient’s response.

After oral administration, lithium is absorbed from the gastrointestinal tract; the rate of peak concentration and plasma fluctuations depend on the formulation. It is not clinically significantly bound to plasma proteins, undergoes no hepatic metabolism and progressively distributes in total body water. Elimination occurs almost exclusively through the kidneys. A substantial proportion of filtered lithium is reabsorbed in the proximal tubule through pathways related to sodium handling: conditions that cause sodium depletion or reduced effective circulating volume therefore tend to enhance its reabsorption and may increase serum lithium levels.

The half life is variable and tends to increase with age, duration of treatment and declining renal function; steady state is reached over several days. Consequently, checking the plasma concentration immediately after a dosage change does not represent steady state exposure. NICE guidelines recommend an initial check about one week after starting treatment and after each dose change, repeated until stabilization; the Italian Summary of Product Characteristics likewise states that assessment should be performed after steady state has been reached.

To compare values obtained at different times correctly, blood sampling must be standardized. In clinical practice, serum lithium is generally measured as a trough concentration at approximately 12 hours after the last dose, unless otherwise specified for the formulation or local protocol. A numerical value without information on the time of the last dose can be misleading. Changes in the daily distribution of the dose can also affect comparisons between serum lithium measurements and should be considered when interpreting an unexpected result.

Lithium also has a particularly close pharmacokinetic relationship with renal function. A reduction in eGFR, an intercurrent illness with fever, vomiting or diarrhea, a marked change in fluid or sodium intake, and the introduction of drugs that interfere with perfusion or tubular reabsorption can turn a previously stable dose into an excessive one. For this reason, prevention of toxicity depends largely on recognizing conditions that alter clearance at an early stage.

Role in bipolar disorder

In bipolar disorder lithium is one of the drugs with the longest record of documented efficacy. International guidelines include it among first line options for acute mania and, especially, among the cornerstone treatments for maintenance therapy. Its antimanic effect is not immediate: in presentations with severe agitation, psychotic symptoms or a need for rapid behavioral control, it may initially be necessary to add an antipsychotic or choose a more rapidly effective strategy, depending on severity and comorbidities.

Maintenance treatment is the setting in which the benefit of lithium is best established. Randomized studies, meta analyses and network meta analyses show a reduced risk of recurrence compared with placebo, with protection against both manic and depressive episodes, although the response profile is not identical in all patients. A previous good response to lithium, clear episodicity with interepisode remission and a classic bipolar presentation have been associated in observational studies with a greater likelihood of benefit, but none of these features constitutes an absolute predictive criterion.

For maintenance treatment in adults, the ISBD/IGSLI task force reached consensus on a standard serum lithium level of 0,60–0,80 mmol/L, with the option of reducing this to 0.40–0.60 mmol/L when response is good but tolerability is poor, or increasing it to 0.80–1.00 mmol/L when response is insufficient and tolerability is good. NICE recommends 0.6–0.8 mmol/L for people starting lithium and suggests considering 0.8–1.0 mmol/L for at least six months when relapse occurs during treatment or subthreshold symptoms with functional impairment persist. These ranges are clinical targets and do not replace the instructions for the specific authorized formulation.

In acute mania some guidelines use higher plasma targets than for maintenance treatment, provided they are compatible with tolerability and renal function. However, a single international range should not automatically be applied to every product: product information may specify different limits and the risk of toxicity increases with exposure. The target level should therefore be defined according to the clinical phase, formulation, applicable recommendations and patient characteristics.

In acute bipolar depression lithium has a clinical rationale and may be appropriate, particularly when it is already indicated for longitudinal stabilization, but evidence for its acute antidepressant efficacy as monotherapy is less robust than the evidence available for maintenance treatment. Modern recommendations should therefore not be summarized by stating that lithium is always the first choice antidepressant treatment for a bipolar depressive episode. Selection depends on response history, predominant polarity, risk of switching, current maintenance therapy and alternatives with evidence specific to the depressive phase.

A potential advantage of lithium in the longitudinal treatment of mood disorders concerns suicide risk. Historical meta analyses of randomized studies and numerous observational studies have suggested a reduction in suicide or self harm; however, a more recent meta analysis restricted to randomized trials judged the evidence inconclusive. A possible antisuicidal effect can therefore appropriately be considered among favorable factors in the overall assessment, but it should not be presented as a certain, universal property independent of control of the mood disorder.

Augmentation in unipolar major depression

In major depressive disorder, lithium is not normally used as initial antidepressant monotherapy. Its most relevant use is as augmentation, meaning addition to an antidepressant when an adequate treatment has produced an insufficient response. Meta analyses of controlled studies support the efficacy of this strategy, historically documented with tricyclic antidepressants and subsequently also with second generation antidepressants.

The decision to introduce lithium should be made within a diagnostic and therapeutic reassessment. Before augmentation, it is necessary to verify that the disorder is genuinely unipolar, that the antidepressant dose and duration are adequate, that adherence is sufficient and that there are no medical conditions, substances or medications that could explain the lack of response. The possibility of bipolar features should also be reassessed because the occurrence of hypomania or mania, or a suggestive history, substantially changes the pharmacological strategy.

NICE states that when lithium is used to augment an antidepressant, therapeutic plasma concentrations are generally at least 0.4 mmol/L and should not exceed 1.0 mmol/L. Lower levels may be appropriate in older patients, taking renal function and greater sensitivity to adverse effects into account. The concentration should nevertheless be interpreted in relation to response, tolerability and individual risk: no serum lithium level guarantees an antidepressant response.

Benefit should be assessed with repeated clinical measures and over a time frame consistent with the therapeutic objective. If no clinically significant benefit emerges, indefinite continuation exposes the patient to monitoring and risks without demonstrated benefit. If augmentation is effective, duration should be individualized according to the number and severity of episodes, relapse risk, tolerability and preferences, while avoiding abrupt discontinuation.

In unipolar depression, lithium may be particularly useful when an augmentation strategy with extensive clinical experience and a mechanism distinct from that of the baseline antidepressant is desired. However, it is not automatically preferable to other augmentation strategies: comparison with second generation antipsychotics, antidepressant combinations, intensified psychotherapy or other options depends on the clinical profile and applicable recommendations.

Serum lithium levels, titration and monitoring

Treatment should begin only after confirming that reliable monitoring can be performed. Before prescribing, NICE recommends documenting weight or BMI, renal function with eGFR, urea and electrolytes, calcium, thyroid function and full blood count; an ECG is indicated in the presence of cardiovascular disease or risk factors. In people of reproductive potential, reproductive planning, possible pregnancy and perinatal risks should be addressed before treatment is started or modified.

Serum lithium should be checked about one week after starting treatment and about one week after each dosage change, followed by frequent measurements until stabilization. During the first year, NICE generally recommends a check every three months. After the first year, the interval can be extended to six months in stably controlled patients, while it remains every three months in older people, those taking interacting treatments, those at risk of renal, thyroid or calcium abnormalities, and those with poor clinical control, poor adherence or higher plasma levels.

Renal function, electrolytes including calcium, thyroid function and weight or BMI should be reassessed at least every six months, or more often when abnormalities, worsening eGFR, endocrine symptoms or conditions that increase risk are present. The rate of eGFR decline is more informative than a single isolated creatinine measurement; persistent deterioration requires reassessment of monitoring frequency, concomitant medications and the benefit to risk balance, with nephrology consultation when appropriate.

Monitoring is not limited to laboratory testing. At every visit, tremor, ataxia, dysarthria, impaired coordination, cognitive changes, polyuria, polydipsia and other signs of possible neurotoxicity, which can occur even when the measured concentration does not appear exceptionally high, should be sought. During chronic treatment, clinical severity may not correlate perfectly with a single serum lithium level, particularly when exposure has increased gradually.

Patient education is an integral part of monitoring. Fluid and sodium intake should remain relatively stable, and situations that can rapidly alter clearance, such as vomiting, diarrhea, fever, profuse sweating, prolonged fasting or major dietary changes, should be recognized. During an acute illness with a risk of dehydration, the patient should contact the physician so that treatment and serum lithium can be reassessed promptly rather than changing the dose independently.

Short and long term adverse effects

The most common adverse effects include nausea, gastrointestinal symptoms, fine tremor, thirst, polyuria and sometimes weight gain. Some are more prominent in the early stages or around plasma peaks and may improve with adjustment of the dose or formulation. In contrast, a new or distinctly coarser tremor, particularly when associated with gait instability, dysarthria or confusion, should raise suspicion of toxicity.

In the kidneys, lithium reduces urinary concentrating ability and can cause nephrogenic diabetes insipidus, clinically characterized by polyuria and polydipsia. During prolonged treatment, chronic tubulointerstitial nephropathy with reduced eGFR has also been described in a minority of patients. Risk increases with duration of exposure, episodes of intoxication and concomitant renal risk factors. Reduced renal function does not automatically require discontinuation: the decision requires comparison of psychiatric benefit, the eGFR trajectory and risk of further damage.

Endocrine effects are well documented. Lithium can inhibit thyroid hormone release and promote goiter, increased TSH and hypothyroidism; risk is greater in the presence of preexisting thyroid vulnerability. Treatable hypothyroidism does not necessarily require lithium discontinuation when psychiatric benefit is substantial. Lithium is also associated with hyperparathyroidism and hypercalcemia, which is why serum calcium is included in periodic monitoring.

Other reported effects include edema, worsening of some dermatoses, electrocardiographic abnormalities and sinus node dysfunction in predisposed individuals. Lithium can unmask or worsen Brugada syndrome, which is also mentioned in Italian product information, and a compatible personal or family history requires particular attention. Patient reported cognitive effects, such as slowing or reduced alertness, should be assessed by distinguishing the contribution of lithium from that of the mood disorder, other medications and hypothyroidism.

Weight gain can occur but is not uniform and may be influenced by thirst, caloric beverages, hypothyroidism, comorbidities and concomitant treatments. Weight monitoring therefore has more than a cosmetic purpose: it allows an unfavorable metabolic trajectory to be identified early and reversible causes to be assessed. Similarly, new onset polyuria should not be dismissed as an inevitable effect because it may indicate a clinically significant defect in urinary concentrating ability.

Drug interactions and factors that alter serum lithium levels

The most important interactions are those that reduce renal elimination of lithium. NSAIDs can increase serum lithium levels, with differences among individual agents; NICE recommends avoiding self medication with NSAIDs in patients treated with lithium. If an anti inflammatory drug is essential, prescribing should be planned and accompanied by closer monitoring of lithium concentrations and renal function.

Thiazide diuretics can markedly increase serum lithium levels through changes in sodium balance and tubular reabsorption. ACE inhibitors and angiotensin II receptor blockers can reduce clearance and precipitate toxicity, particularly in the presence of renal impairment, dehydration or older age. Starting or stopping these drugs therefore requires planned reassessment of serum lithium levels.

An important nonpharmacological interaction is a change in fluid and sodium balance. Abruptly initiated low sodium diets, dehydration, gastroenteritis, fever, physical activity with intense sweating, or conditions that reduce circulating volume can increase lithium reabsorption. Conversely, changes that increase elimination can reduce exposure. The aim is not to impose a high salt intake but to maintain relatively consistent habits and report significant changes.

Combination with serotonergic drugs does not necessarily produce a pharmacokinetic interaction, but it can increase the burden of neurological effects and, with specific combinations, contribute to the risk of serotonin syndrome. Antipsychotics and other psychotropic drugs can also add to tremor, sedation, extrapyramidal symptoms or metabolic abnormalities. Periodic review of the entire treatment regimen is therefore more useful than a simple static list of “prohibited” drugs.

A concentration that has been stable for months does not guarantee that the same dose will remain safe after a clinical change. Every new medication, deterioration in renal function or episode of dehydration should be considered a potential change in exposure. This explains why chronic toxicity is often iatrogenic or facilitated by intercurrent events rather than by intentional overdose.

Acute and chronic toxicity

Lithium toxicity can result from acute ingestion, progressive accumulation during chronic therapy, or an acute overdose in a patient already receiving treatment. These scenarios are not equivalent: in chronic toxicity lithium has had time to distribute into tissues, and important neurological manifestations can occur even at concentrations lower than those observed after a recent acute ingestion. Severity should therefore not be defined by a single laboratory threshold.

Initial symptoms may include nausea, vomiting, diarrhea, worsening tremor and weakness. Neurological progression is characterized by coarse tremor, ataxia, dysarthria, nystagmus, hyperreflexia, confusion and somnolence; in more severe cases, seizures, coma and cardiovascular instability may occur. In a patient taking lithium, the development of new cerebellar signs or altered mental status should be considered potentially toxic until clinically assessed.

Management requires lithium discontinuation and urgent assessment with serial serum lithium measurements, renal function, electrolytes, volume status and ECG when indicated. Rehydration with appropriate fluids is a cornerstone in hypovolemic patients but must be adapted to cardiac and renal status. Extracorporeal treatments are reserved for severe poisoning; EXTRIP recommendations integrate concentration, neurological symptoms, renal function, expected elimination kinetics and other clinical features.

After severe poisoning, serum lithium may rebound because of redistribution from tissues, particularly after hemodialysis, so serial measurements are required. Persistent neurological sequelae have been described in exceptional cases. Prevention remains the most effective safeguard: recognition of interactions, prompt serum lithium measurement during intercurrent illness and clear information about warning symptoms.

Pregnancy, breastfeeding, older adults and kidney disease

In pregnancy the decision is complex because it must balance the risk of maternal relapse against fetal and neonatal risks. Modern cohort studies have confirmed a dose dependent increase in the risk of cardiac malformations after first trimester exposure, but with an absolute risk lower than the most alarming historical estimates. Lithium should therefore be described neither as risk free nor as automatically prohibited in every pregnancy: the decision requires specialist assessment, consideration of alternatives and planned monitoring.

During pregnancy, lithium clearance can vary with physiological changes in glomerular filtration and may decrease again near delivery. A patient who is already stabilized therefore requires an individualized plan with more frequent monitoring and attention to vomiting, dehydration and interacting obstetric medications. The postpartum period carries a high risk of relapse in bipolar disorder, and the strategy for continuing or restarting lithium should be defined before delivery whenever possible.

Lithium passes into breast milk and the neonate has more limited elimination capacity than an adult. Breastfeeding recommendations are not uniform and depend on dose, prematurity, the infant’s clinical status and the feasibility of specialist pediatric monitoring. The issue should therefore not be reduced to a simple yes or no without considering the individual context.

In older adults renal clearance is often reduced, interactions with antihypertensives, diuretics and NSAIDs are more frequent, and neurotoxicity can be mistaken for neurological disease or delirium. Lower doses and conservative plasma targets are generally required, defined according to indication and tolerability. Falls, tremor, cognitive deterioration or worsening balance should prompt reassessment of both serum lithium levels and the entire concomitant medication regimen.

In chronic kidney disease the decision to start or continue lithium requires particular caution. Reduced eGFR increases the risk of accumulation, but discontinuing a highly effective treatment can cause significant relapse. NICE recommends considering psychiatric efficacy, the degree and rate of renal deterioration and, when necessary, nephrology advice. In patients with significant renal impairment, the specific product information may impose restrictions or contraindications that must be observed.

Discontinuation and long term follow up

Lithium should not be stopped abruptly unless there is an urgent clinical need, such as suspected toxicity. For planned discontinuation, NICE recommends gradual dose reduction over at least four weeks and preferably up to three months, even when another antimanic drug has been started. Rapid discontinuation is associated with an increased risk of recurrence, particularly during the first few months.

After discontinuation, early signs of mania and depression should be monitored for at least three months and a relapse prevention strategy maintained. In patients who have had an exceptionally good response for years, the decision to discontinue because of renal or endocrine effects requires particularly careful assessment, because the lost benefit may be substantial and response to reintroduction should not be assumed.

Long term follow up should integrate efficacy and safety. Therapeutic success is not simply a serum lithium value “within range,” but a combination of mood stability, good functioning, tolerability, absence of toxicity and the ability to maintain appropriate monitoring. The lowest dose capable of sustaining clinical benefit with safe exposure is generally preferable to indiscriminately pursuing high plasma values.

Summary table

Bibliography
1. Yatham LN et al. Canadian Network for Mood and Anxiety Treatments (CANMAT) and International Society for Bipolar Disorders (ISBD) 2018 guidelines for the management of patients with bipolar disorder. Bipolar Disorders. 20(2), 2018: 97-170.
2. Fountoulakis KN et al. Lithium treatment of bipolar disorder in adults: a systematic review of randomized trials and meta-analyses. European Neuropsychopharmacology. 54, 2022: 100-115.
3. Nolen WA et al. What is the optimal serum level for lithium in the maintenance treatment of bipolar disorder? A systematic review and recommendations from the ISBD/IGSLI Task Force on treatment with lithium. Bipolar Disorders. 21(5), 2019: 394-409.
4. Nelson JC et al. A systematic review and meta-analysis of lithium augmentation of tricyclic and second generation antidepressants in major depression. Journal of Affective Disorders. 168, 2014: 269-275.
5. McKnight RF et al. Lithium toxicity profile: a systematic review and meta-analysis. The Lancet. 379(9817), 2012: 721-728.
6. Gitlin M. Lithium side effects and toxicity: prevalence and management strategies. International Journal of Bipolar Disorders. 4(1), 2016: 27.
7. Ferensztajn-Rochowiak E et al. Long-Term Lithium Therapy: Side Effects and Interactions. Pharmaceuticals. 16(1), 2023: 74.
8. Haissaguerre M et al. What an endocrinologist should know for patients receiving lithium therapy. Annales d'Endocrinologie. 83(4), 2022: 219-225.
9. Patorno E et al. Lithium Use in Pregnancy and the Risk of Cardiac Malformations. New England Journal of Medicine. 376(23), 2017: 2245-2254.
10. Cipriani A et al. Lithium in the prevention of suicide in mood disorders: updated systematic review and meta-analysis. BMJ. 346, 2013: f3646.
11. Nabi Z et al. Effects of lithium on suicide and suicidal behaviour: a systematic review and meta-analysis of randomised trials. Epidemiology and Psychiatric Sciences. 31, 2022: e65.
12. Decker BS et al. Extracorporeal Treatment for Lithium Poisoning: Systematic Review and Recommendations from the EXTRIP Workgroup. Clinical Journal of the American Society of Nephrology. 10(5), 2015: 875-887.
13. Grandjean EM et al. Lithium: updated human knowledge using an evidence-based approach: part III: clinical safety. CNS Drugs. 23(5), 2009: 397-418.
14. National Institute for Health and Care Excellence. Bipolar disorder: assessment and management. Clinical guideline CG185. NICE, 2014, updated 2025.
15. National Institute for Health and Care Excellence. Depression in adults: treatment and management. NICE guideline NG222. NICE, 2022.
16. Agenzia Italiana del Farmaco. Carbolithium: Riassunto delle caratteristiche del prodotto. Banca Dati Farmaci AIFA, documentazione vigente consultata nel 2026.