Lithium Prophylaxis: Myths and Realities

Lithium is used in the treatment for recurrent bipolar disorder and has helped many patients. However, it is not effective in unipolar and in rapidly cycling affective illness, in long-term use it destroys the thyroid gland and the kidneys, side effects of lithium treatment are frequent and troublesome, and lithium treatment curbs creativity. In addition, lithium intoxications can develop capriciously even with therapeutic doses and serum concentrations; when used in combination with neuroleptic drugs, lithium produces irreversible brain damage; the need for frequent laboratory monitoring makes lithium treatment cumbersome and expensive, and lithium is used too much.

These are widely held views on lithium treatment; some are not complete misconceptions. Since the treatment is administered to many patients, it behooves us to assess the extent of the truth in these statements through analysis of existing evidence or provision of new evidence.

Effect in unipolar illness

Doubt about the efficacy of prophylactic lithium treatment in unipolar illness seems almost exclusively a U.S. phenomenon. In other parts of the world lithium is used with good effect in recurrent unipolar affective illness, and the bulk of the published evidence clearly indicates equal prophylactic efficacy of lithium in unipolar and bipolar illness.

There are some unanswered questions here. Are American psychiatrists unable to distinguish between neurotic or chronic depressions and recurrent depressions with symptom-free intervals? Are they too impatient to wait for the sometimes gradual onset of the prophylactic action of lithium? Or is unipolar illness diagnosed differently on either side of the Atlantic? These questions need answers, for it cannot be without interest whether American unipolar patients are guarded against an ineffective drug or deprived of a valuable treatment.

Effect in rapid cyclers

It is true that lithium treatment often fails in manic-depressive illness with four or more episodes per year, but so do all other treatments, and no comparative trial has yet established superiority of one therapy over another. One may recall that the first observation of relapse-preventive action against depressions was made in a patient with more than 10 episodes per year (1), that in the first systematic study of lithium prophylaxis several patients had many episodes per year and yet responded satisfactorily (2), and that lithium was found highly effective in a patient who for 13 years had suffered from a 48-hour periodic psychosis (3).

These reflections are not meant to argue against trying levothyroxine, carbamazepine, or clorgyline in rapidly cycling patients. They only serve to suggest that one should give the rapid cyclers a chance also with lithium.

Long-term effects on thyroid function

Lithium’s effect on the thyroid gland was first observed in 1968 (4, 5). Recent studies have shown that in a group of lithium-treated patients there is a significant rise in mean serum thyrotropin-stimulating hormone (TSH) and a significant fall in mean serum thyroxine 6–12 months after the start of lithium treatment, but thereafter the values normalize (6, 7).

Hypothyroidism requiring treatment may arise in some patients, and reports about the frequency of this side effect range from 1% to 30%. These figures are, however, of little value when no distinction is made between incidence and prevalence. At the psychiatric hospital in Risskov my colleagues and I have carried out a systematic incidence study (6). In 202 patients who were treated with lithium for up to 6 years, hypothyroidism occurred with an incidence of 2 per 100 years of lithium exposure; the same figure has been reported from Sweden (7). All of our hypothyroid patients responded to supplementary treatment with thyroxine; this does not speak of destruction of the thyroid gland.

Long-term effects on kidney function

Observation in the 1970s of morphological kidney changes in lithium-treated patients generated grave concern among psychiatrists, who asked themselves whether the patients’ mental health was bought at the expense of their kidney function and whether patients given lithium treatment for many years would eventually develop renal failure and die. Many centers decided to subject the matter to systematic scrutiny, and over the last decade the kidney function of more than 800 patients has been examined in longitudinal studies and that of more than 2,700 patients in cross-sectional studies.

The literature dealing with lithium treatment and kidney function has recently been reviewed (8), and the outcome is clear. Lithium treatment does not, even when given for many years, lead to any change of the glomerular filtration rate. After more than 35 years of lithium use in psychiatry and after treatment of large numbers of patients, not a single case of renal failure has been observed that can with any certainty be ascribed to the lithium therapy.

It was the morphological kidney changes that made such an impression in the 1970s, because they were felt to reflect severe kidney damage and a serious prognosis. The evaluation is different today (9, 10). Structural changes may occur in the glomeruli of lithium-treated patients, but they are nonspecific and can be seen also in patients about to start lithium treatment. The morphological changes that are specifically associated with lithium treatment are confined to the distal tubules and collecting ducts, are reversible, and do not signify risk of falling glomerular filtration rate or renal failure.

In recent years references to “the nephrotoxic effect of lithium” have appeared almost routinely in discussions about and in papers dealing with prophylactic alternatives to lithium. It is time that this misleading and anxiety-inducing practice came to an end. Lithium treatment is not nephrotoxic.

The frequency of side effects

Many side effects of lithium treatment are strongly dependent on the treatment intensity and can be avoided or reduced through use of lower doses. In 1979 at the psychiatric hospital in Risskov my colleagues and I reduced lithium doses and serum levels by about 30%. Before 1979 the average dose and serum level were 33 mmol/day and 0.85 mmol/liter, respectively; since 1979 the values have been 23 mmol/day (corresponding to a little less than 900 mg of lithium carbonate) and 0.68 mmol/liter. The frequency and intensity of lithium-induced side effects have been markedly lower after 1979 (11–13) than before (14). There has been no notable reduction of prophylactic efficacy.

Side effects have been studied longitudinally in the patients treated after 1979 (13). Complaints of tremors were presented by 5% of the patients before lithium treatment and by 15% during lithium treatment. The increase was transitory; after a few years of treatment the frequency did not differ significantly from the frequency before treatment. The patients’ body weight rose, on the average, 4 kg during lithium treatment. The increase took place within the first 6–12 months; thereafter the mean body weight remained constant. Complaints of loose stools and urge to defecate rose from 1% before to 6% during lithium treatment; the frequency increased further at serum lithium levels higher than 0.8 mmol/liter.

About one-tenth of the patients presented psychological complaints, which might or might not have been caused by the lithium treatment. These complaints included memory impairment and difficulty concentrating, tiredness, “grayness of life,” and, in a few cases, altered sense of taste and lowered libido or potency.

The lesson to be learned from these observations is that the physician should spend time and effort on adjusting lithium doses and serum concentrations to those levels which give the patient a maximum of prophylactic effect with a minimum of side effects. A serum lithium concentration between 0.5 and 0.8 mmol/liter is appropriate for most patients, but adjustment to values outside this range is necessary for some (15, 16). Even within the recommended range, fine individual adjustment of doses and serum levels may yield benefit; sometimes a change in serum lithium level as small as 0.1 or 0.2 mmol/liter upward or downward can make all the difference for the patient’s quality of life during maintenance treatment.

Effect on creativity

Lithium treatment is given to persons who suffer from recurrent manic-depressive illness, and the crucial question is whether their creativity is affected more when the illness is treated prophylactically with lithium than when it is untreated or treated with neuroleptics and antidepressants. In a study of 24 lithium-treated manic-depressive artists (17), six reported that their creativity was lower during lithium treatment than before, and four of them stopped treatment for this reason (two resumed it later). Six artists felt no change in creativity, and 12 reported that they created more and in some cases better when lithium treatment had brought their illness under control and they could work with more steadiness and better artistic discipline.

Development of intoxication with therapeutic doses and serum levels

The occasional appearance of single-case reports about lithium intoxication in patients being treated with “therapeutic doses” seems to have left the impression that lithium poisoning can develop unexpectedly even when treatment guidelines are followed. Is that in fact so?

My colleagues and I have recently recorded all lithium intoxications that occurred over a 9-year period in the region in and around Aarhus; serum lithium determinations for this region are carried out by the laboratory at the psychiatric hospital in Risskov (18). For patients given lithium treatment in this region, the total exposure time was about 4,900 years. Lithium intoxication developed in 24 patients; none of the intoxications developed without discernible cause.

Fifteen intoxications were caused by deliberate self-poisoning with suicidal intent; no patient died. Perhaps we should impress on our patients that lithium is neither a pleasant nor a reliable self-destructive agent (19). Nine intoxications resulted from neglect of treatment guidelines. Some were caused by continued lithium treatment with unaltered dose during physical illness with fever (four cases) and during lowered intake of food and fluids in a depressive relapse (one case). One patient’s physician gave him lithium doses that were too high; one patient took, in senile confusion, more lithium than prescribed; and two patients took doses that were too high because they felt good on a regimen of lithium and thought they would feel better with more lithium. No intoxication developed as a result of decreasing glomerular filtration rate.

Rather than of capriciousness, these findings speak of a high degree of predictability as far as lithium toxicity is concerned; unintended intoxications can be avoided by adherence to treatment guidelines and precautions.

Risk involved in combining lithium and neuroleptics

There is both clinical and experimental evidence that lithium and neuroleptic drugs can interact adversely. The question is whether, as has been claimed, such interaction typically leads to permanent brain damage, whether the interaction is frequent, and whether it is unavoidable. The evidence indicates that we can answer all three questions negatively.

A recent analysis of the literature led to the conclusion that permanent brain damage was no more frequent in intoxications caused by lithium given together with haloperidol than it was in intoxications caused by lithium given alone (20). Three surveys of a total of 587 patients given combination treatment with lithium and haloperidol failed to find a single instance of adverse interaction (21–23). One-third of the systematically followed Risskov patients were given neuroleptics together with lithium, and yet none showed evidence of adverse interaction (13).

The critical factor seems to be the drug dose. Lithium doses should not be higher than required to produce serum levels below 0.8 mmol/liter in most patients and between 0.8 and 1.0 mmol/liter in a few treatment-resistant ones. Haloperidol administered together with lithium should presumably not be given in doses higher than about 20 mg/day, and other neuroleptics should be given in correspondingly moderate doses. If these rules are followed, there seems to be no reason to abandon a combination therapy that is of considerable value in the treatment of acute mania and in the maintenance treatment of schizoaffective illness.

Need for laboratory monitoring

In our cohort study no patient developed renal failure. Increased urine volume and decreased renal concentrating ability occurred in some patients, but they did not predict deterioration of glomerular filtration rate, for there was none (12). It seems doubtful whether the time, effort, and cost of regular determinations of creatinine clearance, serum creatinine concentration, 24-hour urine volume, and renal concentrating ability are justified any longer. More important for treatment safety are the need to alert patients and relatives to risk situations (physical illness with fever, rigorous dieting, and combined treatment with lithium and diuretics or with lithium and nonsteroidal anti-inflammatory agents) and the need to instruct patients not to take more lithium than prescribed. If unusual signs and symptoms develop, patients should contact the physician, and, if the need arises, laboratory testing should be readily available.

Serum lithium determinations are presumably of sufficient use for the monitoring of lithium treatment to justify their employment at the start of treatment, after dose changes, and when lithium intoxication is suspected. It may further serve a didactic purpose and increase compliance that serum levels are checked at appropriate intervals; the length of these intervals can be individually customized according to the patients’ clinical condition, their understanding of treatment guidelines and precautions, and the need for contact between patient and psychiatrist.

Lithium-induced hypothyroidism may simulate a depressive relapse and may therefore remain undiagnosed by the psychiatrist. Determinations of serum TSH, a sensitive indicator of decreased thyroid function, may help to resolve diagnostic doubts.

The prevalence of lithium use

Estimates have been made of how widely lithium treatment is used, but they have often been based on unreliable methods. We have recently made a detailed survey of lithium use in and around Aarhus (24). The material for analysis was provided by information about patient identity submitted together with blood samples sent to the laboratory in Risskov for serum lithium determination. We could therefore count not only the number of determinations carried out but also the number of patients actually treated.

From the total number of patients given lithium during the years 1986 and 1987, we deducted the incidence of lithium treatment, i.e., the number of patients who started lithium treatment during this period. A point prevalence of 540 in a population of 372,000 was arrived at. In other words, 1.5 of every 1,000 persons in the population had been in lithium treatment at any given time in the region under study during the period under study.

Does this mean that lithium was used too much? Or does it mean that lithium was used too little? These questions are difficult to answer. There were presumably patients who would have been better off if they had not been given this treatment, and there were in all probability patients who might have benefited if they had been given it.

In the Aarhus region, prophylactic lithium treatment has been in fairly general use for more than 15 years. This presumably means that both the phase of skeptical underuse and the phase of enthusiastic overuse have been passed, and it may not be unreasonable to assume that use of lithium by 1.5 persons of every 1,000 in the population is a reasonable estimate of the need for such therapy as used according to present-day indications.

Conclusions

The analysis presented here indicates that the opinions presented at the start of this article are not altogether wrong, for lithium treatment may be cumbersome and expensive, it may produce frequent and troublesome side effects, and it may lead to dangerous intoxications if it is used incorrectly. But we need not do that. Little extra effort is required to follow the guidelines and precautions that ensure maximum efficacy and safety of lithium treatment, and our manic-depressive patients are entitled to just that.