Print Friendly 

William Coryell, MD

Primary Psychiatry. 2002;9(12):47-50

Dr. Coryell is professor in the Department of Psychiatry at the University of Iowa College of Medicine in Iowa City.

Disclosure: This work was supported by an unrestricted educational grant from GlaxoSmithKline.



How does the clinician best manage side effects associated with mood stabilizers used for bipolar affective disorder? Informed management of side effects can improve a drug’s tolerability and therefore increase the likelihood that patients will continue to take the drug most effective for their illness. Placebo-controlled studies have supported the prophylactic use of lithium, carbamazepine, valproate, and lamotrigine for bipolar disorder. This review summarizes the case reports and controlled studies pertinent to the management of side effects for each of these agents.



While the importance of pharmacotherapy in the management of bipolar affective disorder has been established, all of the useful agents have also been associated with side effects. Side effects may be endured, they may lead to treatment noncompliance, or they may result in the replacement of a drug that is effective in controlling illness symptoms with an alternative that is possibly less so. Any of these possibilities will diminish the patient’s quality of life.

In comparison with efforts to demonstrate the efficacy of mood stabilizers in the acute and prophylactic treatment of bipolar affective disorder, the literature concerning the management of side effects is small, largely anecdotal, and rarely reviewed. The following undertakes such a review, with suggestions relevant to each of the mood stabilizers for which placebo-controlled trials have shown efficacy. The side effects discussed are the ones more commonly described and particular emphasis is given to those for which corrective measures other than drug discontinuation have been proposed.


Despite recent shifts in prescribing patterns, the efficacy of lithium in bipolar affective disorder is more thoroughly established than the efficacy of any other medication. Because its use was widespread before the advent of most of the other drugs now being prescribed for bipolar affective disorder, lithium’s side effects have been more fully characterized and a number of strategies for managing them
have emerged.


Nausea may be the first side effect encountered when lithium is introduced, particularly when doses are rapidly increased. Because nausea tends to reflect the rapidity with which plasma levels are increasing,1 loading procedures will increase its likelihood and this may adversely bias the patient’s attitude toward lithium therapy. An obvious solution would therefore be a temporary decrease in dose, followed by a more gradual approach to the targeted plasma level range. Also, nausea often improves when lithium is consistently taken with meals. A slow-release preparation may reduce nausea, but may also increase difficulties with diarrhea.2


A fine tremor, increased by voluntary movement, is often another early development in lithium therapy. It is often tolerated but may be quite distressing for the patient. If the tremor is seen to peak in intensity within several hours of each dose, a sustained-release preparation may be helpful. Also potentially helpful is a reduction in caffeine intake and the elimination of other drugs, such as tricyclic antidepressants, that may produce or worsen tremor. Because maintenance lithium levels at the lower end of the conventional therapeutic range produce less tremor,3,4 doses should be titrated against symptom control to determine whether lower doses provide relief.

When a distressing tremor persists despite dosing changes, modest doses of a β-blocker are indicated. According to a number of reports, propranolol 30–80 mg/day is often effective.5-7 β1 blockers, such as atenolol and metoprolol, have been used for patients at risk for bronchospasm.7,8

Polyuria and Polydipsia

Polydipsia is typically nonprogressive9 but may substantially interfere with sleep, and may lead to significant weight gain if the patient is relieving thirst with sugared drinks. As with tremor, symptoms are more likely at higher doses, and the problem often subsides when careful titration is used to determine the patient’s lowest effective dose.

According to several authors, polyuria is less likely with single daily dosing than with multiple-dose schedules.10-13 At least one study demonstrated that a shift from multiple dosing to single dosing reduced urinary volume,14 but another applied random assignment to alternate dosing schedules and found no difference in urine volume.15

If changes in dose and timing are unhelpful, the addition of a diuretic may produce a paradoxical reduction in urine volume. Hydrochlorothiazide 50 mg/day has been recommended, but amiloride,16 in doses titrated to 10 mg BID, is less likely to produce hypokalemia.

Weight Gain

Weight gain attributed to lithium can be an important issue for long-term compliance and may lead a clinician to consider alternative mood stabilizers. The results of randomized, prophylactic studies are pertinent. In the only comparison of lithium, valproate, and placebo, patients taking valproate, but not those taking lithium, experienced significantly more weight gain than those taking placebo.17 A randomized comparison of lithium and carbamazepine, on the other hand, indicated that carbamazepine resulted in significantly less weight gain.18 Existing evidence does not support a switch from lithium to valproate to limit weight gain.

As an alternative to switching mood stabilizers, clinicians should consider whether coadministered drugs may be contributing to the problem and whether these are necessary. Atypical antipsychotics are often used adjunctively to hasten symptom control in acute mania and then continued out of concern that monotherapy with lithium may be insufficient for prophylaxis. A gradual discontinuation is warranted to determine whether this is so.

Patients should also be asked whether they are drinking high-calorie beverages to relieve thirst or to counteract the metallic taste often experienced with lithium. A regular exercise program should be encouraged as this may enhance the control of both weight and depressive symptoms.19

Cognitive Impairment

Goodwin and Jamison20 found that  the most frequently cited reason for lithium noncompliance was “memory problems.” Clinicians often attribute complaints of mental sluggishness to residual depressive symptoms. However, placebo-controlled studies  have shown that at conventional therapeutic levels, lithium produces similar complaints in normal individuals21 and that deficits are measurable objectively as a slowing in cognitive performance.22

Lithium treatment is well-known to cause hypothyroidism. This is most commonly reflected in elevated thyroid-stimulating hormone (TSH) levels, but approximately 4% of patients on chronic lithium treatment develop abnormally low T3 and T4 levels.23 It is also well-known that depressive symptoms and complaints of sluggishness are typical concomitants of clinical hypothyroidism. Most clinicians who use lithium screen for hypothyroidism routinely and provide thyroid replacement when thyroxin levels are below normal limits.

However, subclinical hypothyroidism may also be a factor in both persistent depression24-26 and cognitive impairment.27,28 A trial of thyroid replacement is therefore warranted when either of these coexist with elevated TSH levels. T4 is commonly given but converging evidence now suggests that T3 provides more benefit in this role. Patients given T3 to augment antidepressant treatment in one study had significantly better outcomes than did those randomized to T4.29 In another study, those patients undergoing thyroid replacement with T4, who were randomized to receive part of their replacement as T3, reported significantly better mood and energy than did those who continued to receive T4 alone.30


As with lithium, many of the side effects associated with valproate are dose-dependent and can be reduced through attention to plasma levels. Bowden and colleagues31 identified 45 mcg/mL as the minimum therapeutic level. Changes in manic symptoms did not correlate with plasma levels above this threshold, but side effects increased markedly when levels exceeded 125 mcg/mL. This increase reflects dose-dependent protein binding and therefore, conditions that affect protein binding conditions may shift the threshold downward. Thus, older patients, especially those with liver disease, and those taking other drugs that compete for protein binding, may experiencea nonlinear increase in side effects beginning at a lower plasma level.


Anticonvulsants that increase γ-aminobutyric acid, as valproate does, tend to cause more sedation than do anticonvulsants, such as lamotrigine, which operate through the modulation of glutamatergic neurotransmission.32 While sedation offers an advantage in the acute control of manic symptoms, it becomes problematic during prophylaxis. Because this complaint does not seem to correlate with plasma levels within the therapeutic range, dose decreasing is often not helpful. If sedation is a problem during maintenance, the clinician should consider whether concomitant drugs, such as antipsychotics or benzodiazepines, can be eliminated at this stage.

Gastrointestinal Symptoms and Weight Gain

The most frequent side effects encountered by patients taking valproate are gastrointestinal in nature.33 Such complaints are less likely with divalproex than with valproic acid.33 If nausea persists with divalproex, a histamine2 blocker may prove helpful.34

Weight gain is at least as likely with valproate as with lithium.17 If weight gain is a prominent concern for a particular patient, the alternative mood stabilizers carbamazepine18,35 and lamotrigine36 should be considered. The continuation of any antipsychotics being given concomitantly should be reconsidered. As noted previously,a regular exercise program can have a number of benefits beyond weight control.


Valproate inhibits urea synthesis and the resulting elevations in serum ammonia levels can have clinical consequences varying from mild fatigue to coma.37 The neurology literature contains many references to this effect, though the psychiatric literature does not. This effect does not appear to be dose dependent and its manifestations—fatigue, cognitive slowing, and sedation—may be misinterpreted as depressive symptoms or even viewed as a positive response.38 Moreover, the tendency of valproate to increase ammonia levels is probably common rather than idiosyncratic.

Hjelm and colleagues39 showed that ammonia levels increased by an average of 101% among five healthy subjects given valproate at doses tapered to 1,500 mg/day over 1 week. Thus, an increase in lethargy after the establishment of valproate therapy warrants a check of ammonia levels. Some have suggested the administration of carnitine 1 g BID as a remedy.40 This may work by altering the effect of valproate on urea synthesis.


Carbamazepine was the first anticonvulsant to be widely used in bipolar affective disorder. There have been no direct comparisons of efficacy between carbamazepine and valproate, and little is known regarding which patients can be expected to respond to one rather than the other. This makes the differing side-effect profiles a relatively important factor as clinicians choose among the anticonvulsants.

Blood Dyscrasias

Carbamazepine has a well-known tendency to lower white cell counts. In perhaps the largest survey of this effect,41 carbamazepine was five times more likely to produce leukopenia than was valproate, though it was not more likely to produce severe leukopenia, defined as a white blood cell (WBC) count <3,000 mm3. The onset of leukopenia is most likely during the first 30 days of treatment and is unlikely after 60 days. The WBC threshold42 below which discontinuation is recommended has varied from 3,000 to 4,000 mm3. Notably, 11 of the cases with moderate leukopenia described by Tohen and colleagues41 continued on carbamazepine without adverse consequences. Lithium is known to stimulate leukocyte production and may serve to correct the leukopenia induced by carbamazepine.43 This combination would be particularly appropriate in cases of incomplete clinical response to carbamazepine monotherapy.


Hyponatremia is a relatively common side effect of carbamazepine and may be symptomatic.44 The patient’s medical regimen should be reviewed when low sodium values are noted. Discontinuation of carbamazepine may be averted by adjusting or eliminating coadministrated diuretics.


Lamotrigine offers advantages over valproate and carbamazepine in terms of such side effects as fatigue, somnolence, and weight gain.32 Weight, on average, has been shown to remain stable or to decrease during lamotrigine treatment.45 Although lamotrigine has only recently come into use as a mood stabilizer, it has been widely used as an antiepileptic and more than 583,000 patient-years of experience have accumulated.46 Lamotrigine also lacks most of the side effects associated with the antidepressants developed for use in major depressive disorder. Several recent placebo-controlled studies have shown lamotrigine to have both acute and prophylactic benefits for bipolar depression,47,48 suggesting that this agent is an alternative to the more traditional antidepressants in the management of bipolar depression.

The most frequent adverse events leading to discontinuation in controlled trials have been skin rashes. Indeed, reports of toxic epidermal necrolysis or Stevens-Johnson syndrome have made some clinicians reluctant to use lamotrigine. An expert panel estimated that rashes occur in approximately 10% of patients given lamotrigine and that those severe enough to result in hospitalization occur in 0.3% of adults and 1% of children.46 Established risk factors for lamotrigine-induced rash are use in childhood, use in combination with valproate, and most importantly, a rapid progression in dose. Recognition of this last risk factor led to revisions in the recommended dose-progression.49 A large-scale comparison of rates before and after the adoption of these revisions shows that the risk for severe rash (associated with systemic disturbances) was seemingly eliminated, though the risk for nonserious rashes remained at 9%.49


Only a few medicines have demonstrated effectiveness in the maintenance of bipolar affective disorder. Because of this, and the fact that most patients with bipolar illness require maintenance for an indefinite period, side effects should be managed with particular care to prevent the unnecessary abandonment of a potentially valuable treatment.

Treatment options fall into four categories. The most obvious category is dose adjustment to titrate symptom prophylaxis against side effects within therapeutic plasma level ranges. An often neglected but frequently rewarding intervention is the trial elimination of concomitant medications which may no longer be necessary. The optimization of thyroid status may improve cognitive slowing or fatigue, particularly during lithium maintenance. Emphasis on a regular exercise program is often preferable to switching medication in the face of weight gain. Finally, modest doses of adjunctive medication have been reported as useful in the management of certain side effects. Most of these apply to lithium side effects, doubtless because lithium has been in use far longer than the alternatives.


1.    Persson G. Lithium side effects in relation to dose and to levels and gradients of lithium in plasma. Acta Psychiatr Scand. 1977;55:208-213.
2.    Persson G. Plasma lithium levels and side effects during administration of a slow release lithium sulphate preparation and lithium carbonate tablets. Acta Psychiatr Scand. 1974;50:174-182.
3.    Maj M, Starace F, Nolfe G, Kemali D. Minimum plasma lithium levels required for effective prophylaxis in DSM-III bipolar disorder. Pharmacopsychiatry. 1986;9:420-423.
4.    Gelenberg AJ, Kane JM, Keller MB, et al. Comparison of standard and low serum levels of lithium for maintenance treatment of bipolar disorder. N Engl J Med. 1989;321:1489-1493.
5.    Kirk L, Baastrup PC, Schou M. Propranolol treatment of lithium-induced tremor [letter]. Lancet. 1973;2:1086-1087.
6.    Lapierre YD. Control of lithium tremor with propranolol. Can Med Assoc J. 1976;114:619-624.
7.    Zubenko GS, Cohen BM, Lipinski JF. Comparison of metoprolol and propranolol in the treatment of lithium tremor. Psychiatry Res. 1984;11:163-164.
8.    Dave M. Treatment of lithium induced tremor with atenolol. Can J Psychiatry. 1989;34:132-133.
9.    Smigan L, Bucht G, von Knorring L, Perris C, Wahlin A. Long-term lithium treatment and renal functions. Neuropsychobiology. 1984;11:33-38.
10.    Plenge P, Mellerup ET, Bolwig TG, et al. Lithium treatment: does the kidney prefer one daily dose instead of two? Acta Psychiatr Scand. 1982;66:121-128.
11.    Bowen RC, Grof P, Grof E. Less frequent lithium administration and lower urine volume. Am J Psychiatry. 1991;148:189-192.
12.    Hetmar O, Brun C, Ladefoged H, Larsen S, Bolwig TG. Long-term effects of lithium on the kidney: functional-morphological correlations. J Psychiatr Res. 1989;23:285-297.
13.    Hetmar O, Povlsen UJ, Ladefoged J, Bolwig TG. Lithium: long-term effects on the kidney. Br J Psychiatry. 1991;158:53-58.
14.    Perry PJ, Dunner FJ, Hahn RL, Tsuang MT, Berg MJ. Lithium kinetics in single daily dosing. Acta Psychiatr Scand. 1981;64:281-294.
15.    O’Donovan C, Hawkes J, Bowen R. Effect of lithium dosing schedule on urinary output. Acta Psychiatr Scand. 1993;87:92-95.
16.    Batlle DC, von Riotte AB, Gaviria M, Grupp M. Amelioration of polyuria by amiloride in patients receiving long-term lithium therapy. N Engl J Med. 1985;312:408-414.
17.    Bowden CL, Calabrese JR, McElroy SL, et al. A randomized, placebo-controlled 12-month trial of divalproex and lithium in treatment of outpatients with bipolar I disorder. Arch Gen Psychiatry. 2000;57:481-489.
18.    Coxhead N, Silverstone T, Cookson J. Carbamazepine versus lithium in the prophylaxis of bipolar affective disorder. Acta Psychiatr Scand. 1992;85:114-118.
19.    Babyak M, Blumenthal JA, Herman S, et al. Exercise treatment for major depression: maintenance of therapeutic benefit at 10 months. Psychosom Med. 2000;62:633-638.
20.    Goodwin FK, Jamison KR. Manic-Depressive Illness. New York, NY: Oxford University Press; 1990.
21.    Judd LL, Hubbard B, Janowsky DS, Huey YH, Attewell PA. The effect of lithium carbonate on affect, mood, and personality of normal subjects. Arch Gen Psychiatry. 1977;34:346-351.
22.    Judd LL, Hubbard B, Janowsky DS, Huey LY, Takahashi KI. The effect of lithium carbonate on the cognitive function of normal subjects. Arch Gen Psychiatry. 1977; 34:355-357.
23.    Amdisen A, Andersen CJ. Lithium treatment and thyroid function: a survey of 237 patients in long-term lithium treatment. Pharmacopsychiatry. 1982;15:149-155.
24.    Frye MA, Denicoff KD, Bryan AL, et al. Association between lower serum free T4 and greater mood instability and depression in lithium-maintained bipolar patients. Am J Psychiatry. 1999;156:1909-1914.
25.    Joffe RT, Levitt AJ. Major depression and subclinical (grade 2) hypothyroidism. Psychoneuroendocrinology. 1992;17:215-221.
26.    Joffe RT, Marriott M. Thyroid hormone levels and recurrence of major depression. Am J Psychiatry. 2000;157:1689-1691.
27.    Prohaska ML, Stern RA, Mason GA, Nevels CT, Prange AJ. Thyroid hormones and lithium-related neuropsychological deficits: A preliminary test of the lithium-thyroid interactive hypothesis. J Int Neuropsychol Soc. 1995;1:134.
28.    Prohaska ML, Stern RA, Nevels CT, Mason GA, Prange AJ. The relationship between thyroid status and neuropsychological performance in psychiatric outpatients maintained on lithium. Neuropsychiatry Neuropsychol Behav Neurol. 1996;9:30-34.
29.    Joffe RT, Singer W. A comparison of triiodothyronine and thyroxine in the potentiation of tricyclic antidepressants. Psychiatr Res. 1990;32:241-251.
30.    Bunevicius R, Kazanavicius G, Zalinkevicius R, Prange AJ Jr. Effects of thyroxine as compared with thyroxine plus triiodothyronine in patients with hypothyroidism. N Engl J Med. 1999;340:424-429.
31.    Bowden CL, Janicak PG, Orsulak P, et al. Relation of serum valproate concentration to response in mania. Am J Psychiatry. 1996;153:765-770.
32.    Ketter TA, Post RM, Theodore WH. Positive and negative psychiatric effects of antiepileptic drugs in patients with seizure disorders. Neurology. 1999;53(5 suppl 2):S53-67.
33.    Zarate CA, Tohen M, Narendran R, et al. The adverse affect profile and efficacy of divalproex sodium compared with valproic acid: a pharmacoepidemiology study. J Clin Psychiatry. 1999;60:232-236.
34.    Stoll AL, Vuckovic A, McElroy SL. Histamine subscript 2-receptor antagonists for the treatment of valproate induced gastrointestinal distress. Ann Clin Psychiatry. 1991;3:301-304.
35.    Placidi GF, Lenzi A, Lazzerini F, Cassano GB, Akiskal HS. The comparative efficacy and safety of carbamazepine versus lithium: a randomized, double-blind 3-year trial in 83 patients.
J Clin Psychiatry. 1986;47:490-494.
36.    Isojarvi JI, Rattya J, Myllyla VV, et al. Valproate, lamotrigine, and insulin-mediated risks in women with epilepsy. Ann Neurol. 1998;43:446-451.
37.    Kifune A, Kubota F, Shibata N, Akata T, Kikuchi S. Valproic acid-induced hyperammonemic encephalopathy with triphasic waves. Epilepsia. 2000;41:909-912.
38.    Eze E, Workman M, Donley B. Hyperammonemia and coma developed by a woman treated with valproic acid for affective disorder. Psychiatr Serv. 1998;49:1358-1359.
39.    Hjelm M, Oberholzer V, Seakins J, Thomas S. Valproate-induced inhibition of urea syntheseis and hyperammonaemia in healthy subjects [letter]. Lancet. 1986;2:859.
40.    Raby WN. Carnitine for valproic acid-induced hyperammonemia [letter]. Am J Psychiatry. 1997;154:1168-1169.
41.    Tohen M, Castillo J, Baldessarini RJ, Zarate C, Kando JC. Blood dyscrasias with carbamazepine and valproate: a pharmacoepidemiological study of 2,228 patients at risk. Am J Psychiatry. 1995;152:413-418.
42.    Joffe RT, Post RM, Roy-Byrne PP, Uhde TW. Hematological effects of carbamazepine in patients with affective illness. Am J Psychiatry. 1985;142:1196-1199.
43.    Brewerton TD. Lithium counteracts carbamazepine-induced leukopenia while increasing its therapeutic effect. Biol Psychiatry. 1986;21:677-685.
44.    Yassa R, Iskandar H, Nastase C, Camille Y. Carbamazepine and hyponatremia in patients with affective disorder. Am J Psychiatry. 1988;145:339-342.
45.    Frye MA, Ketter TA, Kimbrell TA, et al.
A placebo-controlled study of lamotrigine and gabapentin monotherapy in refractory mood disorders. J Clin Psychopharmacol. 2000;20:607-614.
46.    Guberman AH, Besag FM, Brodie MJ, et al. Lamotrigine-associated rash: risk/benefit considerations in adults and children. Epilepsia. 1999;40:985-991.
47.    Calabrese JR, Bowden CL, Sachs GS, Ascher JA, Monaghan E, Rudd GD. A double-blind placebo-controlled study of the lamotrigine monotherapy in outpatients with bipolar I depression. J Clin Psychiatry. 1999;60:79-88.
48.    Calabrese JR, Suppes T, Bowden CL, et al.
A double-blind, placebo-controlled, prophylaxis study of lamotrigine in rapid-cycling bipolar disorder. J Clin Psychiatry. 2000;61:841-850.
49.    Wong IC, Mawer GE, Sander JW. Factors influencing the incidence of lamotrigine-related skin rash. Ann Pharmacother. 1999;33:1037-1042.