• Response will be defined as 50% reduction in MADRS score. [ Time Frame: weekly ] [ Designated as safety issue: Yes ]
  
• Remission will be defined as MADRS ≤ 12. [ Time Frame: weekly ] [ Designated as safety issue: Yes ]
  
• The MADRS will be done at week 0,1,2,3,4,5,6. [ Time Frame: weekly ] [ Designated as safety issue: Yes ]
  

• The Hamilton Depression Rating Scale (HAM -D)-17 item scale, at weeks 0 and 6. [ Time Frame: weeks 0 and 6 ] [ Designated as safety issue: Yes ]
  
• Hamilton Anxiety Rating Scale (HAM-A),at weeks 0 and 6. [ Time Frame: weeks 0 and 6 ] [ Designated as safety issue: No ]
  
• The Young Mania Rating Scale (YMRS), at weeks 0,1,2,3,4,5,6. [ Time Frame: weekly ] [ Designated as safety issue: Yes ]
  
• The Bipolar Depression Rating Scale (BDRS) (42),at weeks 0,1,2,3,4,5,6. [ Time Frame: weekly ] [ Designated as safety issue: Yes ]
  
• Checklist of DSM IV symptoms of mania/ hypomania, with additional questions assessing the presence of mood lability, abnormally high energy, abnormally high libido, and rage. Done at weeks 0,1,2,3,4,5,6. [ Time Frame: weekly ] [ Designated as safety issue: Yes ]
  
• The Beck Suicide Scale (BSS), at weeks 0, 1, 2,3,4,5, 6. [ Time Frame: weekly ] [ Designated as safety issue: Yes ]
  

Introduction: Bipolar disorder is an illness that consists of distinct episodes or "poles" of both major depression and mania (bipolar I) or major depression and hypomania (bipolar II). Both poles of the illness may be fully present simultaneously in what is referred to as a "Mixed episode". The "poles" of Major depression, Mania and Mixed mood are currently defined by descriptive criteria in the psychiatric diagnostic manual, DSMIV.

Estimates of the prevalence of Bipolar disorder are as high as 6% of the population.(1) It is a severe and potentially lethal illness which has a different course and treatment profile to unipolar depression, commonly known simply as "major depression". In 50-60% of cases however, the initial presentation of bipolar disorder is one of major depression, and several episodes of depression may occur before declaration of bipolarity, by virtue of a manic or hypomanic episode (pre-bipolar depression).(2,3) The mainstay of treatment for unipolar major depression is antidepressants. In contrast, antidepressant monotherapy is contraindicated in Bipolar I depression and must be used very cautiously in Bipolar II illness (4) with much concern for an "unacceptable cost /benefit ratio' because of the risk of antidepressant induction of mania and rapid cycling (increase in frequency of episodes as a consequence of antidepressants). (5). Hence treatment of the early depressive episodes of bipolar illness requires recognition of depression specific to bipolar disorder and appropriate choice of medication.

Unfortunately, efforts to clearly distinguish bipolar from unipolar depression cross-sectionally have been mostly disappointing. Most evidence supports the value of family history(6) and there is variable predictive value for specific phenomenological and course descriptors such as hypersomnia, psychomotor retardation, early-onset and psychotic sub types(7,8).

The dilemma then is two-fold:

1. Identifying patients suffering from major depression who are at high risk of bipolar disorder.
2. Choosing appropriate pharmacological treatment.

In choosing pharmacological treatment for a "high risk" group, there is the possibility of causing harm with antidepressants if the true diagnosis is bipolar disorder, and of not treating the depression adequately if using mood stabilizers in what is truly a unipolar illness. This dilemma applies to a large subgroup of those presenting with depression. A recent study in a primary care setting suggested that as many as 21% of patients currently treated with antidepressants for a presumed depressive disorder may in fact have a bipolar diathesis, thus representing a large group that may be improperly treated(9).

The concerns with the use of antidepressants in known bipolar depression have focused on antidepressant - induction of mania, and rapid cycling but these do not always present themselves immediately. However, there may be other subtle and immediate antidepressant treatment effects that make the diagnostic distinction possible and more pressing.(5,10)Indeed, it has been postulated that undetected bipolar depression may explain the apparent poor treatment response and increase in suicidality with the use of antidepressants in the adolescent population, where one might expect an increased prevalence of depression as first episode of bipolar disorder by virtue of earlier age of onset compared to unipolar illness.(11)

Objectives:

The objective of this proposal is to investigate response to acute lithium treatment in subjects who meet the diagnostic criteria for major depression, but are potentially at risk for bipolar disorder, by virtue of family history of bipolarity or completed suicide.

Primary Hypothesis:

Subjects with depression considered at risk for bipolar disorder, will show a significant clinical response to lithium.

Secondary Hypothesis:

1. Subjects with depression considered at risk for bipolar disorder, will show a significant clinical response to paroxetine.
2. Treatment emergent mixed symptoms, in particular agitation, will occur only in the antidepressant treated group.

Study Design:

This is a 6 week double-blind randomized controlled pilot trial of lithium versus paroxetine in subjects with major depression who have either a family history of bipolar disorder or completed suicide. 20 subjects men or women above the age of 18 will be recruited at this centre.

Patients will be randomized to receive either lithium or paroxetine at their first study visit and the dose of drug will be adjusted to find the right one for each patient. Patients will be assessed on a weekly basis thereafter.Data from this pilot study will help in the development of a larger adequately powered study to compare lithium and paroxetine in the treatment of depression in this high risk group.

Patients excluded from the study would include subjects not able to give informed consent,pregnant or breast-feeding women,a current diagnosis of panic disorder, post traumatic stress disorder or psychosis,a history of mania or hypomania,active substance abuse or dependence in the last 6 months, a current depressive episode less than 4 weeks or greater than 12 months in duration, a current or prior adequate trial of lithium or paroxetine,concurrent use of other antidepressants or augmenting agents for the treatment of depression,or a clinically significant medical illness, in particular renal impairment.