Lithium: Fifteen Years Later

This year, we observe the 75th anniversary of introducing lithium into modern psychiatry, attested by the 1949 paper of John Cade in the Medical Journal of Australia [1]. About this event, my editorial in the special Lithium 2010 issue of Neuropsychobiology was titled “Lithium: Sixty Years Thereafter” [2]. Since then, fifteen more years have brought further important information about lithium treatment. This paper makes a narrative review of selected articles published in 2010–2024. They concerned with such topics as lithium’s clinical efficacy, the adverse and beneficial side effects of this ion, the mechanisms of lithium action, and the underutilization of lithium nowadays. The criteria for the selection were at the author’s discretion, following his assessment of the validity of the individual papers.

After three-fourths of a century’s presence in contemporary psychiatry, lithium has still been the recommended first choice medication for the episode prevention in bipolar disorder (BD). This priority has been proposed in all recent guidelines [3‒6] and corroborated in major meta-analyses of the last decade [7, 8]. The prophylactic efficacy of lithium monotherapy exceeds monotherapy with any other mood stabilizer [9]. This may also pertain to pediatric BD where the advantages of lithium over other medications in children and adolescents were established in a randomized trial using placebo [10]. Also, in a sample of three hundred forty pediatric subjects, observed for an average of 10 years, those treated with lithium showed a lower number of suicide attempts, decreased features of depression and aggression, and better psychosocial activity than the subjects given different mood stabilizers [11].

A recent review by the International Society of Bipolar Disorders (ISBD) and the International Group for Study of Lithium-treated Patients (IGSLI) task forces on the characteristics of the prophylactic capability of lithium in BD indicated two positive components such as a periodic course and solid social support and four negative ones such as rapid cycling, more frequent hospitalizations, alcohol, and personality problems [12]. As the components of worse lithium efficacy, the Pharmacogenomics of Bipolar Disorder (PGBD) group pinpointed the presence of anxiety at the onset of the illness, worse functioning, adverse life events, a chronic illness trajectory, the appearance of mixed episodes, suicidality, and migraine [13]. A worse response to lithium is also correlated with a longer duration of untreated disease [14]. Recently, attempts have been made to find early biosignatures of response to lithium (R-LiNK initiative) [15] as well as using lithium binding in the brain for this purpose [16]. Since lithium is a drug with the longest experience of using it as a prophylactic medication, as evidence of such superiority, the observations of patients from Poznan treated with lithium for half a century were published [17, 18].

In 2019, a panel of the ISBD and the IGSLI established the required lithium concentrations for BD prevention. Accordingly, it should be between 0.6 and 0.8 mmol/L, with the prospect of a diminution to 0.4–0.6 mmol/L when a satisfactory clinical outcome and side effects occur or rising to 0.8–1.0 mmol/L when the inadequate outcome but satisfactory tolerability is observed [19]. At the same time, the ISBD Older Adults Bipolar Disorders Task Force established the required prophylactic lithium levels for patients aged 60–79 years as 0.4–0.8 mmol/L and patients aged 80 years and more as 0.4–0.7 mmol/L [20]. Recently, there has also been a concern to the effect of low-dose lithium (serum level below 0.6 mmol/L). Such doses may have pro-cognitive properties opening promising avenues in neurodegeneration and augmentation of affective disorders [21].

In the very first paper ascertaining lithium prophylactic properties, the recurrence-preventive effect of the drug was noted for unipolar depression [22]. It was also confirmed in a meta-analysis published in 2019. In this study, seven randomized controlled trials compared monotherapy with lithium and placebo, while five trials compared lithium and antidepressant drugs. Lithium administration substantially outperformed placebo and was even marginally better than the longitudinal administration of antidepressant medications [23]. However, the augmentation of antidepressant drugs in treatment-resistant depression seems to be the next most important lithium application, following the prophylaxis of BD recurrences [24]. A comprehensive paper on this issue was published in 2013 by German authors led by Bauer et al. [25]. Of interest is also another German study showing excellent results of such an augmentation in geriatric patients [26]. Recently, the issue of lithium augmentation of antidepressant drugs was comprehensively reviewed by Ercis et al. [27].

The side effects of long-term lithium treatment were summarized in a recent article [28]. Among the adverse effects, those related to kidney function after long-term lithium administration have aroused the greatest concern. In 2017, the IGSLI assembly evaluated lithium’s impact on kidneys in 312 bipolar patients coming from twelve collaborating centers, treated with the drug for 8–48 (mean 18) years. Long-term lithium administration was associated with a gradual decline in renal functioning by about 30% more than that due to aging alone. The glomerular filtration rate diminished by 0.7%/year of age and 0.9%/year of treatment [29]. At the same time, a very large nationwide study in Denmark has revealed that patients treated with lithium are no more likely to reach end-stage renal failure than those treated with anticonvulsants valproate or lamotrigine [30]. Another recent study from Sweden showed that the new users of lithium did not differ as to the adverse kidney outcomes from those given valproate although higher lithium concentrations were associated with a higher risk [31]. Also, the guidelines to manage lithium-induced kidney impairment to minimize a decision for drug discontinuation were produced. In patients treated longitudinally, frequent estimation of serum lithium, urea, electrolytes, and creatinine is recommended. The medications that can decrease lithium elimination (e.g., nonsteroidal anti-inflammatory drugs, angiotensin-converting enzyme inhibitors) as well as the acute episodes of lithium toxicity have to be avoided. A referral to a nephrologist should be contemplated if a progressive decline in glomerular filtration rate is observed. Usually, in such cases, the lithium dose is reduced. Sometimes, even a very low lithium concentration of 0.2–0.3 mmol/L may be adequate. A resolution of stopping lithium and substituting it with another mood stabilizer should be made cautiously since the efficacy of another mood-stabilizing drug, particularly in patients hitherto perfectly responding to lithium, may be poor, and the subsequent status of the disease may turn out drug-resistant [32].

The presence of such thyroid abnormalities as elevated thyroid-stimulating hormone, hypothyroidism, and/or goiter when lithium therapy is clinically effective does not necessitate lithium discontinuation. In such cases, a thyroxine supplementation should be performed. In our study, comparing BD patients on lithium for the average of 19 years, with bipolar, gender- and duration of the illness-matched, never receiving lithium, thyroid-stimulating hormone concentration was significantly higher in those given lithium. However, the percentage of hypothyroidism was comparable in these 2 subgroups of patients (24% vs. 18%) and several times higher in female than in male subjects. In patients treated with lithium, the appearance of hypothyroidism commonly took place in the initial years of lithium administration, especially in those having a family history of thyroid abnormalities [33]. There have also been suggestions that lithium influences thyroid immunity; however, in our study, there were no differences in thyroid autoantibodies between lithium-treated and lithium-naive BD patients [34].

Recently, Vandermeulen et al. [35] reviewed the association between lithium and hyperparathyroidism. They found that in lithium-treated patients, the prevalence of hypercalcemia is around 4% compared to 0.5% in a healthy population.

Of interest are recent studies on lithium-induced weight gain which is considered as one of the most cumbersome development of lithium therapy. The recent analysis performed by Gomes-da-Costa et al. [36] showed that the risk of lithium-induced weight gain is lower than previously reported and that some other mood stabilizers may have a higher chance of such a condition. In a case-control study, lithium was associated with not significantly more reports of weight gain than lamotrigine, while with fewer reports than olanzapine, valproate, and quetiapine [37].

On the other hand, several beneficial effects of long-term lithium treatment besides prevention of mood recurrences have been confirmed and evidenced in recent years. The main asset of longitudinal lithium administration may be associated with the prevention of suicide. This was confirmed in meta-analyses published in 2013 [38] and 2017 [39]. In a narrative review, Lewitzka et al. [40] describing more than 20 years of experience with lithium’s suicide prevention pointed out that this effect is not connected with the prevention of mood episodes and constitutes a unique property of the drug. However, there have also been controversial reports. Joanna Moncrieff, the famous critic of pharmacotherapy, challenged the preventive activity of lithium on suicide [41]. A quick reaction of the IGSLI pointed out that the paper of her group contains many methodological inadequacies [42]. Another negative data came from a study by Katz et al. [43], investigating the addition of lithium or placebo to usual care in veteran patients with a recent suicidal attempt and finding no difference in the incidence of suicidal-related events. Nevertheless, in the last years, the anti-suicidal effect of lithium was confirmed by systematic reviews of studies involving great cohorts of patients [44, 45], as well as in a most recent meta-analysis covering thirteen randomized controlled trials and showing odds ratio for lithium as 0.491 [46]. Most researchers reckon that due to the anti-suicidal potency, long-term lithium use should be considered in every patient with mood disorders having suicidal tendencies. The recent 15 years are also marked by studies showing an inverse correlation between a number of suicides and lithium concentration in drinking water. These articles came from Japan [47], Austria [48], the USA [49, 50], Greece [51], Argentina [52], Lithuania [53], and Hungary [54]. Although not confirmed in some countries, the phenomenon may suggest a psychotropic action of trace amounts of lithium.

Probably the second most important beneficial effect of lithium is its neuroprotective and anti-dementia action. Lithium administration may increase the volume of some brain structures which was also observed in the IGSLI paper where lithium treatment was related to the bigger hippocampal volume of bipolar patients [55]. As BD poses an elevated chance of dementia, long-term lithium administration significantly decreases such a risk [56]. The recent study of 26,618 patients (mean age 74 years) of whom 548 were exposed to lithium identified Alzheimer’s disease (AD) in 9.7% and vascular dementia in 2.6% of the cohort. Treatment with lithium correlated with a lower risk for both AD (RR 0.55) and vascular dementia (RR 0.36) [57]. A possible therapeutic effect of lithium in AD and mild cognitive impairment, also in small doses, was suggested [58]. In a recent analysis, similarly to in the case of suicide, higher trace lithium levels were likely connected with a lower risk of dementia [59].

Several favorable effects of lithium in various medical conditions were discovered in recent years. Lithium treatment was associated with a decreased risk of osteoporosis [60], greater bone mineral density, and reduced chance for low bone mass in women with BD [61]. A cardioprotective effect of the drug was suggested [62] as well as a small protection against the risk of cancer [63]. In patients receiving lithium, a lower incidence of asthma was observed [64]. It was also established that lithium treatment does not increase the risk of stroke [65] and that of diabetes [66]. The studies on all-cause mortality in Taiwan showed that lithium was superior among mood stabilizers in decreasing all-cause mortality, suicide-related mortality, and natural mortality [67].

Shortly after the outbreak of the COVID-19 epidemic, a paper appeared showing the antiviral activity of lithium on coronaviruses in animals [68]. Given this and a positive experience with the antiviral effect of lithium in herpes infections [69], hope was expressed by the IGSLI group that the drug can be useful for patients with COVID-19 [70]. However, there is only one randomized clinical trial showing that adding lithium reduces the severity of the COVID-19 infection [71]. However, the study of De Picker et al. [72] using electronic records of 26,554 patients showed that the occurrence of COVID-19 was less in patients having therapeutic lithium concentrations (0.5–1.0 mmol/L) than in subjects having lithium concentrations >0.5 mmol/L. In those with therapeutic lithium levels, the frequency of COVID-19 was lower than in patients treated with valproate.

The last 15 years also brought about many novel data on the mechanisms of lithium. Exactly 15 years ago, the Consortium of Lithium Genetics (ConLiGen) was founded as a product of a joint initiative of the IGSLI and the National Institute of Mental Health. The initial report about this event was published in the lithium-dedicated Neuropsychobiology issue of 2010 [73]. The results of the first genome-wide association study of the ConLiGen appeared in 2016, focusing on the molecular-genetic aspects of lithium’s prophylactic efficacy in BD, involving 2,500 patients from 22 participating centers. A region on chromosome 21, containing four single nucleotide polymorphisms, was found to be associated with lithium efficacy. This sector incorporates two long noncoding RNAs that regulate gene expression in the central nervous system [74]. In the next years, the ConLiGen group evaluated, among others, the connection between preventing by lithium recurrences in BD and polygenic risk score (PRS) for varied psychiatric conditions. The PRS for schizophrenia correlated with worse lithium prophylaxis [75]. This corresponded with our study showing a negative association between the prophylactic efficacy of lithium and a tendency to psychotic symptoms [76]. It was found that PRS for major depressive disorder was also negatively associated with such efficacy of lithium [77]. In another research, the PRS for attention-deficit hyperactivity disorder correlated with a poorer prophylactic effect of lithium, and a PRS for schizophrenia was associated with worse adherence to treatment [78].

Genetic research beyond the ConLiGen group used various methodological approaches. In 2013, a review of studies using the “candidate gene” method was published [79]. Evaluation of forty-five genes related to lithium’s prophylactic efficacy revealed the association with the calcium voltage-gated channel auxiliary subunit gamma 2 (CACNG2) gene, known as stargazin, and the neuregulin 1 (NRG1) gene [80]. The whole-exome sequencing performed on monozygotic twins showing a disparity in lithium’s prophylactic response found dissimilarity in six genes, among them being Disrupted-in-Schizophrenia 1 (DISC-1) [81]. A genome-wide methylomic approach (SeqCapEpi) demonstrated distinct features of methylation in patients showing satisfactory or unsatisfactory effects of lithium prophylaxis [82]. A genome-wide association study of Wolter et al. [83] aimed to find a genome region associated with lithium-induced proliferation of neuronal progenitor cells. The location on chromosome 3p21.1 was identified, containing the gene G protein nucleolar 3 (GNL3).

In recent decades, novel methods for studying the mechanisms of lithium were introduced. Among in vitro studies contrasting the characteristics of cells collected from so-called lithium responders (LR) and lithium nonresponders (LNR), the models of induced pluripotent stem cells (iPSCs) and the lymphoblastoid cell lines have been most frequent. The studies of iPSC initiated by the PGBD group [84] demonstrated differences in hyperexcitability of iPSC-derived neurons between LR and LNR. Recently, Khayachi et al. [85] disclosed the mechanisms of restoring neuronal hyperactivity in LR as associated with factors that determine cell growth and proliferation. In the iPSC model, lower expression of the lymphoid enhancer-binding factor 1 (LEF1) gene, which regulates neuronal activity, in LNR was revealed [86]. The dissimilarities concerning the genes associated with neurogenesis between LR and LNR were also obtained by the method of lymphoblastoid cell lines [87].

The immunomodulatory and anti-inflammatory mechanisms of lithium action have been frequently implied in recent years. This is related to the hypothesized role of “low-grade inflammation” in BD. The ConLiGen group examined the prophylactic efficacy of lithium in relation to human leukocyte antigen (HLA) genes. An association of a good response to lithium with HLA genes connected with lower inflammation, and an association of a poorer response with HLA genes related to higher inflammation was revealed [88]. Using lymphoblastoid cell lines, the differences in 335 mRNA and 77 microRNA, the majority of them connected to the immune system, between LR and LNR were identified [89]. These findings were confirmed by showing the differences in immunoglobulin-related genes between lymphoblastoid cells from patients with BD and healthy individuals, as well as between LR and LNR [90]. Recently, a ConLiGen study evaluated a polygenic score (PGS) for immune system genes and its association with the prophylactic effect of lithium. An increased representation of genes associated with cell adhesion and intercellular communication was found, while only a moderate association with genes of the immune system [91]. The therapeutic mechanism of lithium involving immunomodulatory and anti-inflammatory effects has been recently discussed in review articles [92, 93].

In the final part of this review, it would be necessary to mention a concern over the underutilization of lithium despite the favorable data on this drug obtained in many studies. It started in 2018, when the American specialist in BD, Robert Post, regretted greatly the underutilization of lithium in the USA. He indicated the multiple benefits of lithium and pointed out the excessive worry of lithium’s adverse effects [94]. A huge decrease in lithium use in the USA was evidenced in the article published 2 years later [95]. Also in 2018, the author of this review in a paper titled “Challenging the negative perception of lithium and optimizing its long-term administration” reminded the dominance of lithium over other mood stabilizers, the advantages of longitudinal lithium treatment as well as the options of for handling the adverse effects [96]. In a paper published in 2022, the issue of an anomalous association between lithium data and lithium use was emphasized [97]. The 2020 editorial of the journal Bipolar Disorders, titled “Make lithium great again!,” with the Chief Editor of the journal, Gin Malhi, as the first author, appealed to an enhanced exploitation of the therapeutic potential of lithium and wider utilization of the drug [98]. The demands for advanced teaching about lithium application by clinicians and patients have continued as it was supposedly the best strategy for overcoming a decrease of lithium use [99]. A recent editorial in the journal Bipolar Disorders brings a claim by Malhi and Bauer that lithium should be more promoted as the drug of first choice for BD prophylaxis [100].

In the 15 years that passed after the Neuropsychobiology lithium issue, the drug has been confirmed as the most important for recurrence prevention in BD, also in adolescents, and a significant option for augmentation of antidepressants in treatment-resistant depression. Novel information was provided for the possible management of lithium’s adverse side effects, especially those kidney related. Also, the beneficial lithium impacts, of which anti-suicidal and anti-dementia seem the most important, were corroborated. It was also found that lithium reduces all-cause mortality and the morbidity of many conditions, e.g., osteoporosis. Among studies on lithium mechanisms, most of the results of neurobiological (mostly genetic) findings may be related to lithium response and some (e.g., immunomodulatory) to the pathogenesis of BD. However, despite these favorable data about lithium, the use of the drug has shown a tendency to decrease. Therefore, there is an urgent need to better educate both patients and physicians about the most appropriate candidates for lithium prophylaxis, the advantages of lithium use, the management of its possible side effects as well as to start lithium in good nominees in the early phase of the illness. This could make more patients the beneficiary of this drug. The appeals to lithium education and its more widespread use have been on the increase since 2018.

The author has no conflicts of interest to declare.

This study was not supported by any sponsor or funder.

Janusz K. Rybakowski is the sole author.