To the uncritical mind, it appears as if Young & Hammond ^{Reference Young and Hammond1} have made a case for more use of lithium in mood disorders than is currently the trend. They partly based their argument on the meta-analysis by Smith et al. ^{Reference Smith, Cornelius, Warnock, Bell and Young2} A close perusal of the meta-analysis, however, revealed that the case made by Young & Hammond for lithium is one-sided, unbalanced and may be misleading. Even though the study by Smith et al stated that lithium remains the medication with the strongest evidence base, we believe that its declining use may be due to incontrovertible evidence of adverse effects. For example, in the meta-analysis by Smith et al, when withdrawals for any reason and withdrawals for adverse events were analysed, there were more withdrawals with lithium compared with lamotrigine, valproate semisodium and olanzapine. Even in terms of efficacy, the choice of lithium remains arguable. For example, when relapses due to depression were analysed, Smith et al found that there were more relapses with lithium than with lamotrigine and valproate semisodium. In terms of manic episode, there were more relapses with lithium compared with olanzapine, and in terms of any mood episode, there were more relapses with lithium than valproate semisodium and olanzapine.

We do not advocate for any particular medication but we strongly feel that for this type of medication advocacy, authors should attempt to provide a balanced rather than one-sided argument. It is also patronising to partly ascribe the declining use of lithium to poor training of psychiatrists rather than acknowledge the fact that psychiatrists may actually base their choices on individual patient criteria as well as the profile of medications within the wide array of available agents.