How to distinguish and manage bipolar disorder

12 Mar 08

1. How common is bipolar disorder and what features should trigger us to consider the diagnosis? Are there any early features we might miss?

The best estimates of 12-month prevalence of DSM-IV bipolar disorder are 0.6% (1% lifetime prevalence) for type 1, and 0.8% (lifetime prevalence 1.1%) for type 2.

Bipolar disorder should certainly be considered in anyone with recurrent episodes of depressive disorder. Current opinion is that it is worth asking about previous episodes of elation before prescribing antidepressants to anyone who does not have a known history of unipolar depressive disorder. Retrospectively, it can be easy to miss even clear-cut episodes of mood elevation:

• that were not recognised or treated at the time

• where the mood disturbance was mixed or irritable rather than classically elated

• where the patient themselves strongly attributes the symptoms to life events.

2. How easy is it to distinguish between bipolar 1 and 2 without clear longitudinal data? What specific features enable you to make the diagnosis of a manic episode?

Subtyping bipolar disorder on the basis of the history can be challenging and it is always best to get corroborative information from an informant.

Nevertheless, when using reliable diagnostic criteria such as DSM-IV, it’s possible to make a confident diagnosis in most cases.

For type 1, the patient must have experienced an episode that meets criteria for mania. For type 2, the patient must have experienced an episode that meets the criteria for hypomania.

3. Some patients are diagnosed as having ‘rapid cycling’ BPD. How is this defined and are there any factors that make rapid cycling more likely? Does this have implications for choice of drug therapy?

Rapid cycling is defined in DSM-IV as at least four discrete episodes of mood disorder in a 12-month period. Patients with bipolar disorder may have many more frequent episodes than this – at one extreme, their mood may swing severely within days or weeks.

The presence of rapid cycling does have implications for drug therapy. NICE guidelines recommend avoiding antidepressants. Long-term mood stabilisation can be difficult and will often require combinations of agents. Indeed, NICE recommends combination lithium and valproate as first-line therapy.

4. How does bipolar disorder relate to the diagnosis of schizoaffective disorder?

Although our current diagnostic classifications of psychotic disorders have substantial clinical utility, they are imperfect and many patients do not fit neatly into one diagnosis.

There is frequently overlap of cross-sectional symptom profiles between schizophrenia and psychotic mood episodes – which is why a long-term perspective must be taken. In schizoaffective disorder, there must be severe psychotic symptoms occurring outside a mood episode.

5. Do we understand why bipolar disorder can follow head injury? Who is most at risk and does the disease follow a different course in this situation?

An association between head injury and subsequent bipolar disorder has long been reported, although recent epidemiological studies do not demonstrate a very clear or strong relationship. Any aetiological mechanism remains speculative and the specific clinical features of bipolar disorders occurring after head injury are unknown.

6. What is the inheritance pattern of bipolar disorder? Does it vary between the types?

Bipolar disorder is strongly familial and genetic epidemiological studies have identified a few polymorphisms that may increase risk. Current thinking is that any genetic predisposition is most likely to be to a bipolar spectrum with the actual phenotype representing the interaction between polygenetic inheritance and the environment of that particular individual. A great deal of current effort is therefore being put into trying to define the endophenotypic markers that are inherited.

7. What is the response to psychotherapy and should all patients be referred to a psychologist?

None of the main structured psychotherapies – group psychoeducation, family-focused therapy, interpersonal and social rhythm therapy or cognitive behaviour therapy – have a convincing evidence base in bipolar disorder.

It would not be justified to refer all patients to a psychologist, but development of psychosocial therapies for bipolar disorder is an active area of research.

NICE guidelines recommend that structured psychotherapies including psychoeducation and mood monitoring should be offered to patients who still have depressive symptoms after an acute episode. Bear in mind that long-term studies show 32% will be left with depressive.

8. Is lithium still the mood-stabilising agent of choice and are there any tips for monitoring?

Lithium remains the mood stabilising agent with the best evidence base and most of us would consider it the first-line drug treatment of choice – at least in type 1.

It is usually best to keep the level in the range 0.6-0.8mmol/l although 0.4-0.6mmol/l or even lower or 0.8-1.2mmol/l can be used for some patients. Once established, levels should be checked every three to six months and renal and thyroid function checked at least annually.

The decision to start long-term lithium – two years or more – should be based on the best available estimate of the risk of relapse. This may be sufficiently high in, for example, a patient with a severe first manic episode with a strong family history who is very keen to reduce their risk of relapse.

9. If an anticonvulsant drug is used instead of lithium, which agent do you choose, what doses do you aim for and are blood levels important?

The two anticonvulsants with the best current evidence are sodium valproate – mainly in the form of valproate semisodium – and lamotrigine.

Lamotrigine appears to be most effective in preventing depressive episodes, whereas the evidence for valproate does not suggest a specific effect on either pole of the illness.

The target dose for lamotrigine is usually 100-200mg. Some patients need a lower dose – 50-100mg – and I rarely see any additional benefit above 200mg. Doses need to be reduced by 50% if lamotrigine is prescribed in combination with valproate. I do not find blood levels informative.

For treatment of acute mania, the dose of valproate may need to be as high as 2000mg.

For long-term valproate treatment, I would try to achieve a dose of 1,250mg of valproate semisodium – or 1,500mg of other valproate preparations – reducing by steps of 250mg if the patient cannot tolerate that dose. At lower doses it can be worth checking the blood level – less that 50µg/l is probably less effective.

10. What’s your advice about medication choices in young fertile females?

This group is becoming difficult for prescribing, because of potential adverse effects on both fetus (with valproate, lithium and lamotrigine) and on the patient themselves – such as the association between valproate and polycystic ovaries.

Nevertheless, the illness may be severe in these patients and it is important to use the full range of treatments, but with care.

I would advise the patient to take precautions against pregnancy and to avoid valproate if possible. I would withdraw long-term medication if the patient wishes to become – or becomes – pregnant.

In milder disorders, it may be better to avoid long-term treatment and to use short-term courses of antimanic or antidepressant agents.

11. What is your drug of choice in mania if a major tranquilliser is needed?

Unlike schizophrenia – where there is a reasonable evidence base for the conventional or typical antipsychotics – the older drugs were never evaluated properly in acute mania. So we know much more about the newer atypical agents in mania.

Most atypical antipsychotics have now been evaluated. My choice would normally be either risperidone 1-8mg, or olanzapine 2.5-20mg, depending on the patient.

Risperidone is less sedating and less likely to cause metabolic syndrome and weight gain, but there is better longer-term evidence for olanzapine.

I also use other agents, including both conventional agents (chlorpromazine and sulpiride) and other atypical antipsychotics (aripiprazole, amisulpride and quetiapine).

12 If depression is the major feature, what is the best way to manage it? How do you manage those who are manic but then rapidly plunge into severe depression, and those with depression who then become rapidly manic on antidepressants?

We now know that the major cause of long-term symptoms and disability in bipolar disorder is depression (see figure). Treatment of depressive symptoms certainly takes most of my efforts in clinic.

Historically, bipolar depression was treated the same as unipolar depression but, increasingly, the treatment of the two conditions is diverging. Antidepressants, for example, should not be used without adjunctive antimanic therapy in both bipolar 1 and 2 disorder. Or, if the patient does not want combination therapy, they should be advised of the risks of mood destabilisation and how to deal with early signs of mania.

NICE also recommends that antidepressants should not be used in patients with recently unstable mood or a past history of antidepressant-induced mood destabilisation.

Furthermore, they should be discontinued early after remission – in contrast to the recommendations in unipolar disorder, where antidepressants should be continued for six months.

There is definitely a trend away from the use of antidepressants in bipolar depression. This is probably due to the limited evidence of efficacy, concern about mood destabilisation and the emergence and marketing of new therapies including quetiapine – with its increasingly impressive long term and short term evidence – and lamotrigine.

Avoiding post-manic depression is another reason for choosing the newer atypical drugs, such as olanzapine and quetiapine, or mood stabilisers such as valproate. Both olanzapine and quetiapine are antidepressants as well as being antimanic, particularly quetiapine.

13 What advice do you give about prognosis? What if the episode occurs in a young person either after antidepressant use or illicit drug use? Is it still considered true that the disease gets worse with each decade?

I try to estimate the risk of relapse for an individual patient and also to give them some idea of the likely beneficial effect of treatment.

On average, after one severe episode, the annual risk of relapse is around 10-20% . This rises on the basis of number of episodes, so that after five or more episodes the patient’s annual risk of relapse will be around 50%.

A four-year follow-up of first episode patients found that nearly all recovered from the acute episode of mania at two years, but that only 36% recovered full premorbid function.

Some 40% of people had a recurrent manic (20%) or depressive (20%) episode within two years of recovering from the first episode.

If the first episode is clearly caused by antidepressants or drugs, these figures can be less easily applied – nevertheless, there is often sufficient uncertainty about the relationship to make the above figures a useful and prudent default.

Professor John Geddes is honorary consultant psychiatrist at the Oxfordshire Mental Healthcare NHS Trust and professor of epidemiological psychiatry at University of Oxford, UK

Competing interests None declared