Prophylactic therapy with lithium in elderly patients with Unipolar Major Depression.
Wilkinson D, Holmes C, Woolford J, Stammers S, North J.;
Commented by , 19 Aug 2002
Aim of the study
To compare the relapse rate of elderly depressed patients taking low dose lithium plus antidepressant, with those taking antidepressant alone.
Method
50 cognitively intact psychiatric in-patients over 65 (mean age 75.8 years), recovering from DSM-IV major depression, were randomised in a double blind study to receive additional lithium carbonate or placebo. All patients were followed up for 2 years for evidence of relapse.
Patients with major depression were identified by the admitting psychiatrist and rated with the Montgomery Asberg Depression Rating Scale (MADRS). Those scoring >25 on the MADRS were followed up weekly; following recovery (as indicated by the clinical team) and prior to discharge, subjects were screened again with the MADRS and Mini-Mental State Examination (MMSE).
Subjects scoring <13 on MADRS and >23 on MMSE, who were not on lithium, but were to be maintained on antidepressants, were approached for consent for inclusion to the study. Consenting subjects were randomised to receive either lithium (200 mg initially, titrated to serum level 0.3 – 0.7 mmol/l) or placebo, in addition to current antidepressants.
Lithium levels were monitored by a clinician blind to patient details; any change in dose was followed by an equivalent change in the next consentive patient in the placebo group.
Subjects requiring change in antidepressant, admission for ECT, or score >13 on MADRS were considered relapsed and withdrawn from the study.
Results
1 patient withdrew consent, leaving 49 subjects (32 female, 17 male). Mean MMSE score was 28.4, mean MADRS score 4.7 (s.d. 3.3). Mean lithium dose was 348 mg/day, with a mean serum lithium of 0.43 mmol/l.
There was no significant difference in age, gender or baseline MADRS and MMSE in placebo and treatment groups.
At both 6 months and 2 year follow up, there were no significant differences in MADRS scores compared to baseline in either lithium or placebo groups. However, at 6 months, 4 patients had relapsed, all of whom were taking placebo; at 2 years, 9 patients had relapsed, 8 of whom were taking placebo.
These numbers – although small – reach statistical significance on chi-squared tests at probability levels of P = 0.05 and P = 0.01 respectively.
12 cases had discontinued from the study at 2 years, either due to poor compliance (5) or signs of possible lithium toxicity (7 – 3 of whom were on placebo!) A further 4 subjects died of unrelated causes.
Discussion
Long term addition of low dose lithium to antidepressants used in common clinical practice may help to prevent relapse of major depression in elderly subjects.
The strength of this study is that a clinical population was used – ie., depressed patients that clinicians are used to seeing in clinical practice. An obvious weakness is the small numbers, the authors describe the study as a “pilot study”.
This finding is consistent with work published by Wilson et al in 1995 in the British Journal of Psychiatry, with a larger scale study in younger adults reported in the September 2000 edition of the American Journal of Psychiatry (Bauer et al).
Depression in older adults is a severe illness with high morbidity and mortality. Prophylactic treatment that prevents relapse in a high risk group – those recovered from an episode of major depression – is worth considering.
It is noteworthy that in this study, there was no statistical difference between the frequencies of side effects in the lithium-treated group than in the placebo-treated group. If lithium augmentation is to be used in clinical populations, clearly it has to be monitored closely.