There are suggestions that even the Greeks recognised bipolar disorder and that from as early as the second century spring waters that were alkaline and as a result likely to be high in lithium salts were known to be of use in the treatment of overactive states such as mania.15^{16 and 17} The Greeks almost certainly did not recognise manic-depressive illness. ¹

Lithium itself was isolated first by August Arfwedson in 1817. It was named lithium because it was found in stone – lithos being the Greek for ‘stone’. During the 1850s alkaline compounds such as lithium developed a reputation for treating rheumatic disorders and gout by interfering with the precipitation of uric acid in the blood and joints, and lithium was available in many countries through the 1970s for the treatment of rheumatism.

In the 1850s, mania and melancholia according to some were part of the same family of diseases as gout, and this led to the use of lithium for these conditions also. In 1880, Carl Lange in Copenhagen claimed that it had a role in preventing episodes of periodic depression. William Hammond in New York found the same thing.

Despite these discoveries, lithium slipped out of use for mood disorders and had to be rediscovered in 1949. In part this was because older world views that connected gout to manic-depressive illness vanished, and in part it was because of lithium’s side effects – increased urine flow, tremor of the hands and difficulties with memory or concentration. Later, in the 1940s, when used as part of a salt-restriction diet in the USA, lithium was linked to cardiac problems, which led the Food and Drug Administration to ban its use.

In 1949, following observations that lithium had a tranquillising effect on laboratory animals, John Cade, in Australia, gave it to manic, depressive and schizophrenic patients. He noted that it was particularly beneficial in mania. Cade’s observations were followed up by Mogens Schou in Denmark, who confirmed in clinical trials that lithium was beneficial in patients with mania. This led to its subsequent spread for use in the treatment of mania.

The adoption of lithium by the psychiatric community, however, was slow and has remained patchy for several reasons. One is that it can have serious side effects, so that blood lithium levels have to be determined regularly to ensure that its side effects do not outweigh its benefits. Second, lithium as an elemental compound is widely available and therefore no drug company stands to make much money out of it. It has certainly not been marketed as aggressively as other compounds. For 50 years, awareness of its usefulness depended largely on the efforts of Mogens Schou. And third, even for the treatment of mania it took second place to the antipsychotics.

But in the 1960s, studies from the UK and Denmark appeared supporting Lange’s 1880 claim that lithium might be useful in the prevention of recurrent episodes of mania or depression. These claims for a prophylactic effect caused a storm of controversy that, in fact, may have helped the spread of lithium. ¹ One of the arguments of critics was that the results that showed people doing well on lithium and poorly off it might simply be the result of a withdrawal syndrome. This argument was dismissed by lithium’s supporters at the time but it now seems that there is indeed a dependence syndrome, although there appear to be benefits from lithium beyond those of avoiding withdrawal.

Lithium was not available in Japan during the 1960s. It was this that led to an interest in trying carbamazepine in mania and to the discovery that carbamazepine could produce useful inter-episode effects. The effects of valproate on mood were similarly discovered in the 1960s in France, where lithium use never became widespread.^{418 and 19}

In the 1980s, the example of valproate and carbamazepine led to the suggestion that anticonvulsants might help mood disorders in much the same way as they helped convulsive disorders – by reducing kindling. The notion was that each episode of a mood disorder kindled a further episode, in the same way that each epileptic fit increases the vulnerability to the next fit. This hypothesis led on to the systematic testing of every new anticonvulsant that has emerged on the market to see whether it might do something useful. Even electroconvulsive therapy (ECT), it was argued, increased seizure thresholds, making further fits less likely – but little mention was made of lithium.

This idea led to the concept of a mood-stabiliser, a concept first applied to oestrogen and progesterone and later clozapine and cannabis before being picked up by the marketing department at Abbott to promote valproate. In 1995 valproate was launched in the USA as Depakote, and as a mood-stabiliser. There was no evidence that it stabilised moods. If Abbott had claimed Depakote was prophylactic the US Food and Drug Administration would probably have sued them but they could claim it was a mood-stabiliser because no one knew exactly what that term meant. But it suggested prophylaxis – an ability to ward off future episodes. It suggested that Depakote, and all the other drugs now called mood-stabilisers, were new forms of lithium.

There are two ways in which a drug might act as a mood-stabiliser. One would be to reduce kindling, in which case all anticonvulsants should help – but they do not appear to, as is outlined below. Lithium, furthermore, is not anticonvulsant. If mood-stabilisers worked by reducing kindling, the takers of these drugs should not notice anything useful about them other than that they reduced the frequency of episodes of mood disorder, in much the same way that patients on anticonvulsants for epilepsy do not talk about anything useful the drug does for them – they keep a record of whether they are having more or fewer fits.

The other way a mood-stabiliser might work, however, would be by virtue of each drug doing something potentially useful. For instance, valproate is sedative, gabapentin is anxiolytic, carbamazepine has anti-irritability or anti-impulsivity effects, as has lithium, and lamotrigine produces a sense of well-being. If they all act in different ways, conceivably they should suit different patients and, if the drug was helping, patients should be able to say ‘This helps me because it does X or Y or Z’. At present, however, there is no interest in pursuing research in this area, which would be much less consistent with pharmaceutical company interests.

In summary, there are good grounds for saying that, except perhaps for lithium, there is no such thing as a mood-stabiliser. Indeed there is some evidence that, despite the availability of so many more ‘mood-stabilising’ drugs, those patients with bipolar disorder are in fact doing worse now than they were doing 100 years ago. ²⁰ If the various different drugs do not correct an abnormality, then in fact they provide another physiological stressor to an already vulnerable system and are likely in the long run to destabilise and make things worse rather than better. There is therefore a considerable premium on ensuring that people are on a drug that suitsthem, and not just on a mood-stabiliser because that’s what you do for people who have bipolar disorder. Calling something a stabiliser doesn’t make it one.

Table 7.1 list the drugs referred to as mood-stabilisers.

Lithium affects such a large number of physiological processes that 50 years after its introduction there is still no consensus on what its key physiological effects are. The surprise is that it acts so widely throughout the body and yet has relatively specific clinical effects. At present lithium is used in the treatment of manic states. It is sometimes used to treat depression, in conjunction with other antidepressants, as part of a strategy called lithium augmentation. It is most commonly used, however, to prevent recurrent episodes of mania or depression.

Since the early 1960s there has been a clear body of evidence pointing to a role for lithium in the prevention of episodes of mania and depression in bipolar affective disorders. Many individuals who have been treated in hospital for mania are maintained on lithium for years or decades to prevent recurrences in what is known to be a recurrent disorder. The evidence that lithium prevents recurrences is much better than the evidence for anything else.

Furthermore, lithium has also been linked to a lower rate of suicide in bipolar disorders than other treatments. In part this may stem from the fact that compliance with lithium may indicate someone who is generally more responsible and concerned about their condition, and thus at lower risk of suicide anyway. The same argument should apply to valproate and carbamazepine, however, but suicides in patients maintained on lithium seem lower than in these other two patient groups.

There is a considerable amount of evidence indicating a role for lithium in recurrent depression. The current wisdom is that lithium is indicated if there are as many as two episodes per year or three episodes of depression over the course of 2 years. The efficacy of lithium, however, seems to fall off once there are more than four episodes of a depressive disorder a year.

The traditional wisdom had been that it was necessary to start prophylactic lithium after one manic episode, but now any patient with a manic episode is likely to be advised that they need a mood-stabiliser. At the opposite end of the spectrum, lithium does not seem to help in what are called rapidly cycling mood disorders, where there are four or more episodes of a mood disorder per year. Overall, because of the withdrawal effects, there are estimates that patients have to stay on treatment for at least 2.5 years before they are likely to have had fewer episodes than they would have had had they not started lithium. ²¹