See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/14636839

Psychotherapy for Bipolar Disorder

Article  in  The British Journal of Psychiatry · December 1995

DOI: 10.1192/bjp.167.5.581 · Source: PubMed

CITATIONS READS
116 3,882

1 author:

Jan Scott
Newcastle University
522 PUBLICATIONS   21,092 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Burden of care View project

R-Link View project

All content following this page was uploaded by Jan Scott on 22 May 2014.

The user has requested enhancement of the downloaded file.

Psychotherapy for bipolar disorder
J Scott

The British Journal of Psychiatry 1995 167: 581-588

Access the most recent version at doi:10.1192/bjp.167.5.581

References
Article cited in:
http://bjp.rcpsych.org/cgi/content/abstract/167/5/581#otherarticles
Reprints/ To obtain reprints or permission to reproduce material from this paper, please write
permissions to permissions@rcpsych.ac.uk

You can respond http://bjp.rcpsych.org/cgi/eletter-submit/167/5/581

to this article at
Email alerting Receive free email alerts when new articles cite this article - sign up in the box at the
service top right corner of the article or click here

Downloaded bjp.rcpsych.org on July 15, 2011

from Published by The Royal College of Psychiatrists

To subscribe to The British Journal of Psychiatry go to:

http://bjp.rcpsych.org/subscriptions/
British
Journalof Psychiatry
(1995),
167,581—588 Review Article

Psychotherapy
forBipolar
Disorder
JAN SCOTT

Background.Psychosocialfactorsmaycontribute25—30%
to the outcomevariancein bipolar
disorders.Sufferershave identifiedbenefitsfrom psychotherapy,but biologicalmodelsand
treatmentsdominatethe researchagenda.The authorreviewsresearchon psychosocialissues
and interventionsin this disorder.
Method.Researchon adjustmentto the disorder,interpersonal
stressorsandobstaclesto
treatment compliancewere locatedby computerisedsearchesand the author's knowledge
of the literature.All publishedoutcomestudiesof psychosocialinterventionsin bipolardisorder
are reviewed.
Results.There is an inadequatedatabaseon psychosocialfactorsassociatedwith onset and
maintenanceof bipolardisorder.While the outcome studiesavailableare methodologically
inadequate,the accumulatedevidencesuggeststhat psychosocialinterventionsmay have
significantbenefits for bipolarsufferersand their families.
Conclusions.Giventhe significantassociatedmorbidityand mortality, there is a clear need
for more systematicclinicalmanagementthat addressespsychosocialas well as biological
aspectsof bipolardisorder.The author identifiesappropriateresearchstrategiesto improve
knowledgeof effective psychosocialinterventions.

It is estimated that an adult developing bipolar
disorder (BD) in his/her mid-20s effectively loses 9
years of life, 12 years of normal health and 14 years
of work activity. In addition, the suicide-related
mortality and the psychosocial consequences for
‘¿significant
others' identify BD as a significantpublic
health problem (Prien & Potter, 1990). The firstline
treatment for BD remains pharmacotherapy, and the
advent of lithium and other drugs have undoubtedly
improved the quality of life for many individuals.
However, Joyce (1992) noted that, even under
optimal research conditions, prophylaxis will protect
fewer than 50°lo of patients with BD against further
episodes. Given the increasing interest in the use of
psychotherapy in treatment-resistantdepressive and
schizophrenic disorders, it seems surprising that such
initiatives have not been applied more systematically
to individuals with BD. This paper highlights the
reasons why such approaches may have been
ignored, emphasises why they should now be used,
and identifies potential avenues for future research.
Why have psychologicalapproachesbeen
overlooked?
There appear to be three reasons why clinicians have
been reluctant to employ psychosocial interventions:
firstly, aetiological models which highlight the strong
genetic and biological correlates of BD have
dominated the researchagenda; secondly, it was long
believed that patients with BD made a full inter
episode recovery; and thirdly, psychoanalysts have
historically expressed greater ambivalence about the
suitability of psychotherapy for manic—depressive
patients compared with patients with other severe
disorders. The latter is probably the most important
influence. Fromm-Reichmann wrote that in com
parison with individuals with schizophrenia, bipolar
patients were poor candidates for psychotherapy
becausethey lackedintrospection,weretoo dependent
and were likely to discover and then play on the
therapist's ‘¿Achilles'
heel'. Yalom suggested that the
inclusion of a bipolar patient in a therapy group was
“¿oneof the worst calamities―that could occur.
Although others argued strongly for the importance
of psychological treatments (e.g. Benson, 1975),
the relative lack of empirical support for such
developments (no large-scale, randomised, control
trial has ever been published) meant that clinicians
received little encouragement or advice on how to
incorporate such approaches into treatment.
The discouraging views on psychotherapy with
bipolar patients were mainly published in the pre
lithium era. As such, reticence about trying to cure
acute mania with a talking therapy is easily under
stood. However, the holistic approach advocated in
modern practice encourages the integration of
biomedical and psychosocial models of disorder. The
treatment of acute mania rightly focuses initially on
pharmacotherapy, but when the mental state is
stabilised then the devastating impact of the episode
on the individual and his/her family clearly needs
addressing. Even if the patient was functioning well
premorbidly,or apparentlymakes a full inter-episode

581
582 SCOTT

recovery, he/she will need help in coming to terms Kahn (1990) highlights the special problem of ‘¿dual
with having a chronic and recurrent disorder. Also, vulnerability' in individuals with early-onset BD. He
Goodwin & Jamison (1990) have argued that as argues that mood instability or other prodromal
establishing ‘¿control'
of biological factors is essential symptoms preceding onset of the disorder may
to the effective management of BD, it is vital to negatively influence interactions with people at
understand and overcome psychological barriersto home. A vicious cycle develops where the suboptimal
compliance with pharmacotherapy. environment created then adversely affects the
individual's personality development which further
Psychosocialissues in BD damages interpersonal interactions even before the
first illness episode occurs. Goodwin & Jamison
There are a number of problems related to BD and (1990) also argue that early-onset BD may arrest
compliance where psychosocial therapies could have or interrupt the completion of ‘¿developmental
a role but have so far been underused. tasks' that normally lead to independence and
leaving home.
Problems related to the disorder
Premorbid personality and coping skills may predict Loss
an individual's reaction to the diagnosis. Other Individuals with BD may experience significant
psychosocial problems in BD may relate to actual distress or relapse if concrete or abstract losses are
or anticipated losses, or the nature and severity of not acknowledged and addressed. Financial and
the illness and its impact on relationships.
employment problems are cited by 70% of patients
Adjustment problems
and their partners as the most frequent long-term
difficulties (Targum et al, 1981). In the first year after
A comprehensive review of premorbid personality by a manic episode, Harrow et al (1990) reported that
Goodwin & Jamison (1990) suggests that individuals 23°lo of patients were continuously unemployed, and
with BD are more similar to, than different from, 36°lo showed a clear decline from their premorbid
‘¿normal'
controls. Although remitted bipolar patients level of functioning at work. Loss of self-esteem may
show lower rates of personality disorder (23°lo) than result from the loss of status. Relationships may be
remitted unipolar patients (35°lo),adjustment after lost because of irrevocabledamage done by aberrant
discharge is significantly worse, with 60% of BD behaviour during a manic episode. These losses are
cases exhibiting functional impairment (Harrow et often accompanied by feelings of guilt. Overall, the
al,1990).Persistent affective symptoms accountfor apparent lack of control over life undermines any
about half the cases of poor adjustment. belief in self-efficacy and may lead to demoralisation,
The reasons why others with BD show such deficits particularly if recurrencesoccur when the patient is
are less certain as robust premorbid predictors of complying with medication (Kahn, 1990).
adjustment are lacking. What is clear is that calm As well as real current losses, anticipated losses
acceptance of the diagnosis and full compliance may also lead the patient to give up hope for the
with treatment are exceptional. Goodwin & Jamison future. Detection of hopelessness is important as it
(1990) state that on discovering that the illness is is a key variablein determiningwhethersuicidalideas
chronic, recurrent and potentially life-threatening, are acted upon. Inquiring about the meaning of the
predictable reactions are denial, anger, ambivalence illness often reveals that patients now view themselves
and anxiety. All these responses can have adverse asdefective (Rush,1988; Goodwin& Jamison, 1990)
effects if they become protracted. Joyce (1992) noted and express anxieties about the potential loss of
that readmnissioncan be predicted by illnessbehaviour, actual or hoped for relationships, or ambivalence
with those who fail to recognise or respond to about the advisability of having children. Younger
evolving symptoms or who are less accepting of drugs people become more hopeless as they gradually
having a worse outcome. Resentment and frustration realise that they may never achieve career or other
may impair relationships with the family, social goals, or attain autonomy.
network and professionals trying to offer treatment. Goodwin & Jamison (1990) also highlight losses
Bipolar patients with high levels of anxiety often relatedto treatment, but stipulate that realisticlosses
use inappropriate strategies to try to avoid relapses, must be distinguished from unrealistic losses (where
such as excessive self-monitoring and extreme the disorder or the treatment are inappropriately
restrictions on their lifestyle. The perceived stigma blamed for all past and currentdifficulties). Realistic
of the diagnosis may also adversely affect self-image losses may include reduced energy, productivity
and lead to social avoidance. and sexual activity. The loss of creativity associated
 PSYCHOTHERAPY FOR BIPOLAR
with ‘¿highs'
has negative consequences for some
individuals.
Interpersonal relationships
Unlike the extensive literature on unipolar depressive
disorders, there are less data on the effect of BD on
marital and family relationships, or the effects of
these relationships on the prognosis of the disorder.
Frank et al(1981) found that marital adjustment in
couples where one spouse suffered from BD was
similar to that in matched control couples where both
spouses were mentally healthy. Other studies report
less favourable outcomes, with significantly higher
divorce and separation rates and expressed conflict
in the marriages of bipolar patients compared with
unipolar patients and community controls. Brodie
& Leff (1971) reported that divorce in bipolar
patients (which occurred in 57% of their sample) was
always instigated after the first manic episode.
Bipolar patients are also reported to have fewer
confiding relationships than control subjects. Targum
et al (1981) noted that 53°lo of healthy spouses said
that they would not have married their bipolar
partner, and 47% said that they would not have had
children had they known the disorder would occur.
Patients tended to underestimate the impact of the
disorder on their relationships. Overall, healthy
spouses regard the problems associated with the
disorder as more severe and the effects of treatment
as more beneficial than their bipolar partners.
Bipolar individuals and their healthy spouses agree
that depression is easier to cope with and accept than
mania as the spouse tends to be sympathetic and
perceive the disorder as beyond the patient's control
(Targum et al, 1981). The threat of violence and poor
judgement, particularlywith respectto interpersonal
interactions and financial extravagance, tend to
dominate concerns about mania. Healthy spouses
often regard the behaviour of someone with
hypomania as deliberate and spiteful. Interestingly,
Hooley et al (1987) demonstrated that marital
adjustment in spouses of patients who suffer more
florid manic episodes is actually better than in
spouses of patients with less severe symptoms. This
may be because these extreme types of behaviour are
more readily accepted as uncontrollable (Goodwin
& Jamison, 1990).
The literature on the parenting skills of bipolar
patients is sparse. Suggestions that parent—child
bonding may be disruptedby lack of consistent care,
or that mothers with BD are less attentive to their
children's needs, come from small, mainly descriptive
studies and it is difficult to extrapolate to other
samples.
DISORDER 583
The influence of interpersonal relationships on BD
outcome must also be considered. In a study of 23
bipolar patients, Miklowitz et al (1988) found that
intrafamilial levels of expressed emotion (EE) and
affective style(AS)predict thelikelihood ofrelapse
over 9 months' follow-up. Social adjustment after
discharge was also predicted by AS profile. The most
striking finding in this study was a relapse rate of
94% if either EE level was high or AS profile was
negative regardless of treatment regime, medication
compliance, baseline symptoms, illness history or
demography. If the AS profile was benign and EE
level low, the relapse rate was only 17°lo.High relapse
rates and symptom exacerbation despite adequate
lithium prophylaxis have also been noted in bipolar
patients reporting stress related to marital disharmony
or experiencing other interpersonal events.
Problems relating to compliance
Data on medicationcompliancein bipolarcases relate
mainly to lithium. Information on carbamazepine
and other drugs is generally lacking, although non
adherence rates for carbamazepine (38%) tend to be
lower than rates for lithium treatment (51%) (Goodwin
& Jamison, 1990). Up to 75% of relapses in bipolar
disorder may be associated with non-compliance.
Prevalence and patterns of non-compliance
Reported non-compliance rates for lithium prophy
laxis vary between 20—50%,with about one in five
patients failing to comply despite a good therapeutic
outcome. Goodwin & Jamison (1990) make a plea
for more research, commenting that, unlike non
response (which has been extensively examined), non
compliance should be reversible through good
clinical management and a therapeutic relationship
which facilitates the discussion of adherence problems.
Complianceis rarelyall or nothing. Fullcompliance
and total non-compliance are the most overt types
of behaviour, but intermittent and late patterns are
reported. The prevalence of intermittent compliance
is high: 47% of bipolar patients discontinue lithium
against medical advice on at least one occasion and
34% discontinue on two or more occasions. Some
individuals describe a cyclical pattern of strict
compliance immediately after an illness episode,
followed by reducing compliance, leading to non
compliance if they remain symptom free. This
behaviour pattern is reinforced by the fact that side
effects disappear early, often leading to the patient
feeling better, while the reappearance of symptoms
is delayed and not always associated with non
compliance in the patient's mind (Rush, 1988).
584 SCOTT
Late compliance (where early drug refusal is later
replaced by adherence) is of interest as it highlights
the critical need to identify and tackle denial. It
appears that some individuals initially reject both the
diagnosis and the medication. With time, evidence
of disorder and the negative consequences of
untreatedepisodes lead to the development of insight
and gradual acceptance of the rationale for medication
(Goodwin & Jamison, 1990).
Risk factors for non-compliance
Several studies identify that lithium compliance rates
are increased in those with a stable social network,
those who perceive the symptoms as severe and
treatment as effective, and those with obsessional
personality traits (Goodwin & Jamison, 1990). The
most common factors associated with lithium non
compliance are younger age, male gender, experience
of fewer illness episodes and previous history of
treatment non-adherence. Non-compliance rates are
particularly high during the first year of lithium
treatment and in those who have persistent elevated
mood, a history of grandiosity or who complain of
missing ‘¿highs'.
The latter is noteworthy as patient
surveys suggest that fear of depression is a stronger
motivating factor for compliance than fear of mania.
Drug side-effects undoubtedly account for a
significant proportion of cases of non-compliance,
although psychiatrists may rate side-effects as a more
important cause of non-compliance than patients
(Jamison & Akiskal, 1983). Clinicians and sufferers
also differ in their views of which side-effects are
most problematic and which lead to non-adherence
(Jamison et al, 1979). It is reportedthat 75°lo of side
effects regarded by psychiatrists as most important
are somatic symptoms, while 80°loof side-effects
regarded by patients as most worrying are cognitive
changes such as mental confusion and memory
problems (Jamison&Akiskal, 1983).One explanation
for this disparity is that clinicians and patients
disagree about whether a particular feature is a side
effect or a symptom. However, a review of patient
reports of somatic side-effects found that although
certain complications (such as excessive thirst) occur
more frequently, the side-effects they found least
acceptable and most likely to lead to non-compliance
were weight gain and tremor (Goodwin & Jamison,
1990). These findings clearly have implications
for how doctors and patients communicate about
treatment.
Only a small literature exists on individual attitudes
towards lithium treatment (Cochran & Gitlin, 1988;
Rush, 1988; Peet & Harvey, 1991). Recurringthemes
are a greater risk of non-compliance in individuals
who dislike having their moods ‘¿controlled' by
medication and see the disorder and the need to
receive long-term pharmacotherapy as a personal
weakness. They often express the view that if they
simply tried harder then relapses would not occur.
In a study of 48 out-patients, Cochran & Gitlin
(1988) demonstrated that both individual attitudes
and social influences (namely what significant
others expect an individual to do) modify lithium
compliance. Aetiological theories and treatment
advice ofrelatives who suffer frommentaldisorder
strongly influence patient beliefs, and the quality of
the doctor—patientrelationship also significantly
affects compliance. If the patient perceived the
psychiatristas ambivalent about the treatmentor the
patient was not motivated to do as expected, he/she
was less likely to comply. A control trial of 60
lithium clinic attenders also demonstrated that those
offered an educational programmeshowed improved
knowledge and more favourable attitudes towards
pharmacotherapyand better medication compliance
than those receiving standard treatments alone (Peet
& Harvey, 1991).
Outcome studies
There are a number of descriptive articles on the use
of psychosocial interventions in BD, but few
publications address outcome, and empirical treat
ment studies are rare. The available research is
reviewed; none was undertaken in Britain.
Individual
therapy
Anecdotal reports of the benefits of individual
therapy for bipolar patients can be found in the
psychoanalytic and cognitive-behavioural literature.
The largest case series was published by Benson
(1975) who described a 41-month, open, follow-up
study. Twenty-four out of 31 patients (21 women)
who received psychotherapy in addition to lithium
maintenance treatment were reported to have a good
clinical outcome.
Cochran's (1984) study of the impact of six
sessions of individual cognitive therapy on lithium
compliance and clinical outcome in bipolar out
patients is the only one to include a control group.
Twenty-eight patients with BD who were newly
referred to a lithium clinic were randomly assigned
to cognitive therapy or ‘¿treatment as usual'. On
subjective and observer ratings (including serum
lithium levels),compliance was significantly better
at6-weekand6-monthfollow-up intheintervention
group. Only three subjects assigned to cognitive
therapy (21°lo) as opposed to eight (57%) control
 PSYCHOTHERAPY FOR BIPOLAR DISORDER 585
subjects discontinued lithium against medical advice, Finally, Wulsin et al(1988) described a long-term
and admission rates were significantly lower in the group run monthly at a CMHC for 22 bipolar out
cognitive therapy group. patients (12 women) over 4 years. Unlike the other
studies, the group focused on interpersonal relation
ships, and the prescribing of lithium was done solely
Couples and group therapy
outside this setting. Reductions in admission were
Davenport and colleagues (1977) described the use reported, but the drop-out rate was 55°lo.
of psychodynamic ‘¿couples group therapy' (n = 12)
and retrospectively contrasted this approach with
Family therapy
lithium clinicattendance (n=11) and community
mental health centre (CMHC) follow-up (n = 42). Fitzgerald (1972) noted his clinical impressions of the
Individuals were not randomly allocated to the benefits of family therapy for a consecutive series
different treatments, but all those included in the of 25 bipolar patients. Others describe general
study had been admitted for mania 2—10years benefits from this approach, but more systematic
previously and had an intact marriage at the time data are provided by Glick and colleagues (1994).
of discharge. At follow-up, the ‘¿couples group Ina cross-national studyof24in-patients withsevere
therapy' patients werefunctioning significantlybetter affective illnesses, patients who received individual
in terms of social and family adjustment and and family psychoeducation showed better resolution
reported no readmissions or marital failures. The of the index episode and better global outcome
CMHC group had the worst outcome (16 12—18months after discharge.
readmissions; 10 marital failures; 3 suicides), but the A further large-scale, randomised, controlled trial
differences between these patients and lithium clinic using an in-patient family intervention(IFI) has been
attenders were non-significant. Treatment modality undertaken by the above research group (Spencer et
was the only predictor of outcome identified. al, 1988; Clarkin et a!, 1990). This study looked at
No controlled trials of group therapy (combined outcome 18 months after discharge in 169 in-patients
with lithium treatment) exist, but four open studies who received six sessions of IFI during their hospital
have been published. Shakir et al (1979) and later stay. The results for the 50 cases of affective
Vollunar et a! (1981) reported on a therapy group for disorders were reported separately (Clarkin et a!,
15 lithium-responsive, bipolar patients (13 men) which 1990). The data on 21 bipolar subjects (14 women)
ran for a number of years. Shakir and colleagues within this group represent the only randomised
identified significant ‘¿beforeand after' changes in control study of psychological treatments undertaken
patient functioning in the 2 years before and 2 years in BD. Twelve subjects were allocated to IFI and nine
after the introduction of group therapy. Before to the control intervention (the usual in-patient
therapy, ten patients had a history of poor compliance treatmentprogramme).Drop-out ratesand treatment
and admissions, the group spent an average of 16 after discharge did not differ significantly between
weeks a year in hospital, and only five individuals the two groups. The immediate and long-term
were in regular employment. After about 51 weeks' outcome data demonstrated that, in comparison to
attendance ata ‘¿Yalom-style'
group,onlythree indi all other groups, female bipolar IFI patients showed
viduals were admitted over the next 2 years, the significantly better social, family, leisure, work and
average lengthofin-patient staywas3weeksperyear, role performance and significantly improved family
serum lithium levels and compliance rates were attitudes to treatment. Although the gains made by
improvedand ten patientswerein continuousemploy female bipolar IFI subjects diminished with time, the
ment. Volkmar and colleagues wrote that the shared results held even when other variables were controlled
experience of BD and lithium treatment enhanced the for. Female bipolar patients benefited significantly
therapy process, but highlightedthat it was not poss from IFI, whereas male bipolar patients and unipolar
ible to distinguish the specificeffects of psychotherapy patients showed either no benefit or, in some cases,
from the non-specific benefits of close follow-up. a negative effect.
Kripke & Robinson (1985) describe a long-term Prien& Potter (1990) pointoutthatthere aretwo
out-patient support group for 14 bipolar patients populations of bipolar patients to be considered
(13men),eightof whom werestill attending the when providing family therapy: older bipolar
group 12 years later. Anecdotal evaluation suggested patients with a spouse (and children); and young
that problem-solving strategies were better received adults with early-onset BD who live with their nuclear
than dynamic analysis. Perceived benefits were family. The latter group may benefit from an
reduced rates of admission and enhanced social and adaptation of behavioural family therapy (BFT)
economic functioning. previously advocated in schizophrenia by Falboon. A
586 SCOTT
pilot study of eight patients treated with lithium and schizophrenia
about20 yearsago.Work on schizo
BET suggested that the approach was well received phrenia has since demonstrated an interaction
by sufferers and their families (Miklowitz eta!, 1988) between biological vulnerability and psychosocial
and, over 9 months, relapse rates in this group (13°lo)
dimensions, and the need for adjunctive psycho
were significantly lower than those in a similar group therapies is now accepted. Prien & Potter (1990) note
of 23 patients who received lithium alone (70%). that BD may disrupt the patient's family environment,
reduce ability to cope with stress, impair social
Comment adjustment and lead to deficits similar to the negative
symptoms of defect states. Life events and intra
Most of the studies reviewed are unsophisticated or familial stress have also been implicated in early
inadequate in a number of ways: only a small number relapses. On the basis of these data, there are
of recognised research toolsor definedoutcome implications for clinical practiceand future research.
measures were used, and only two studies (Cochran,
1984; Clarkin eta!, 1990)randomly allocated patients
Clinical implications
to either a psychosocial or control treatment. These
are the most comprehensive studies available, also Many clinicians employ psychosocial strategies in
employing more clearly defined treatment modalities the management of BD. While flexibility of style
(CT and IFI). However, the relatively small sample and techniques is often required to cope with
sizes render the statistical power low and make fluctuating moods and other psychopathology, there
definitive statements about outcome unwise. is some consistency in the descriptions of beneficial
Some tentative conclusions can be drawn from the approaches (Rush, 1988; Goodwin & Jamison, 1990),
researchers' observations. Individual therapy clearly and most clinicians reiterate the fmdings of the
improved knowledge about BD and its treatmentand National Institute of Mental Health collaborative
allowedsufferers to explorebeliefs aboutthese study on depression that systematic clinical manage
issues in detail. Group process benefited from the ment in combination with pharmacotherapy is a
homogeneousnatureof thesampleand,provided simple and effective strategy. Organising an agenda
that individuals were engaged in the therapy, for follow-up appointments ensures coverage of all
admissions did not unduly disrupt the proceedings. topics. Assume that compliance will become an issue
Sharing knowledge about the disorder helped those for all patients at some point and create an
who denied or underestimated problems to gain atmosphere in which ambivalence or obstacles can
insight andawareness intotheir difficulties. Couples be anticipated, discussed and simple behavioural
group therapy was helpful in educating spouses about techniquesintroduced (such as ‘¿pairing' tablet-taking
the disorder, allowed exploration of patients' own with a routine activity) to facilitate adherence.
attitudes and reduced stress. Similar benefits accrued Giventheimportance ofotherpeople's attitudes,
from family therapy in both female in-patients (IFI) and research evidence thatsufferers underestimate
and younger out-patients (BET). Older men may the impactof BD, then extending the psycho
do well in individual or group formats. Possible educational approachisappropriate. Franketa!
explanations are that the families of male bipolar (1985) reported that after patients and their families
patients may be more critical than the families of attended a 1-day educational workshop, there were
female patients, or that men with BD are hyper no reported cases of treatment non-compliance.
sensitive to interpersonal stimuli and find family Handouts and videos can also be used to supplement
therapy more stressful (Clarkin et a!, 1990). sessions (Peet & Harvey, 1991). Sufferers and their
relatives may also engage in simple diary-keeping to
identify early warning symptoms of relapse, or
Conclusions
record response to medication so that decisions
For many decades, the dominant research agenda in regarding changes can be made more objectively.
bipolar disorder has been biological. The impact of Identifying ‘¿high-risk situations' for relapse and
drugs on acute symptoms and relapse rates has been developing a hierarchyof coping responses(including
emphasised so strongly that the role of other how to access mental health services) are other
therapies has been ignored. However, there are limits possible uses of such data.
to the efficacy of pharmacotherapy and, although Some research suggests that the first year after the
treatment adherence is a recognised issue, little is onset of BD is a crucial time for patients in terms
known of the psychological barriers to compliance of adjusting to the disorder, developing insight
(Rush, 1988). The paucity of psychosocial research and complying with treatment, systematic clinical
in BD has many parallels with the situation regarding management at this stage may be a simple strategy
 PSYCHOTHERAPY FOR BIPOLAR DISORDER 587
for improving outcome. Lastly, although clear Blocheta!(1994) suggestthatpsychosocial factors
guidelines for BD are not available, the use of contribute about 25—30%to the prognostic variance
cognitive, brief dynamic, interpersonal, couples or in BD. Psychosocial variables that affect compliance
family therapy should at least be considered more with medication will also share a small proportion
often. of the variance attributed to biological factors.
Giventhatpsychosocial aspects make a significant
contribution to outcome, the opportunity for
Research implications
empirical research and the introduction of systematic
There are a number of obvious areas for future clinical interventions should not be ignored. The
research. evidence in this review suggests that BD sufferers and
A greater understanding of individual and their carers would both welcome and benefit from
environmental vulnerability factors that influence such initiatives.
onset or outcome of BD episodes is required. In
unipolar depression there is evidence that cognitive References
dysfunction and problem-solving deficits influence
the course of the disorder. No similar literature exists BENSON, R. (1975) The forgotten treatment modality in bipolar
illness:
psychotherapy.
Diseases
of theNervousSystem,36,
for BD. If certain dysfunctional attitudes were 634—638.
associated with persistence of symptoms, then it BLOCH, S., HAFNER, J., H@s.a@i, E., et al (1994) The Family in
might be possible to develop non-drug strategies to Clinical Psychiatry, pp. 92-108. Oxford: Oxford Medical.
modify these beliefs. The work on AS and EE also BRODIE, H. & Lass, M. (1971) Bipolar depression - a comparative
study of patient characteristics. American Journal of Psychiatry,
needs to be extended to larger patient samples. If the 127,1086—1090.
evidence confirms that drugs do not protect the CIAIUCIN, J., GLICK, G., H,@s, G., et al (1990) A randomized
individual against the adverse effects of high EE, it clinical trial of in-patient family intervention. V: results for
clearly has implications for the use of adjunctive affective disorder. Journal of Affective Disordeis, 1*, 17-28.
COCHRAN,S. (1984) Preventing medical non-compliance in the out
family therapy techniques. Lastly, while many life patient treatment of bipolar affective disorder. Journal of
events are unavoidable, their occurrence signals a Consultingand ClinicalPsychology,52, 873—878.
period of increased risk of relapse in patients with & GITLIN,M. (1988) Attitudinal correlatesof lithium
BD and offers an opportunity for integrated compliance in bipolar affective disorders. Journal of Nervous
and Mental Diseases, 176, 457-464.
biopsychosocial research. DAVENPORT, Y., EBERT, M., Aoi@im, M., et a! (1977) Couples
The problems of medication compliance need to therapy as an adjunct to lithium maintenance of the manic
be examined. Only 1—2% of the papers written on patient. Journal of Orthopsychiatry, 47, 495-502.
lithium explore this, and even less is known about FrI-i.oaxdu.o,R. (1972) Mania as the message: treatment with family
therapy and lithium carbonate. American Journal of Psycho
adherence to other drugs. Frank eta! (1985) suggest therapy, 26, 547—535.
that researchers have a duty to identify and account F@a, E., PRIEN,
R.,Kupsea,
D., etal(l985)Implications
ofnon
for cases of non-compliance in treatment studies so compliance on research in affective disorders. Psychopharma
thatwe canimproveourunderstanding. More work cology Bulletin, 21, 37—42.
T@aow.i, S., GERSHON,E., et al (1981) A comparison of
on attitudes towards BD and its treatment should be non-patient with bipolar-well spouse couples. American Journal
undertaken. Family attitude studies were mainly of Psychiatry,13*, 764-768.
undertaken in the USA some time ago and therefore GLICK, I., Buan, L., OKON0GI, K., et al (1994) Effectiveness in
need to be repeated. psychiatriccare. III: psychoeducation and outcome for patients
with major affective disorder and their families. British Journal
Randomised control trials are required to establish of Psychiatry,164, 104-106.
the short- and long-term benefits of psychotherapy GOODWIN, F. & JAMISON, K. (1990) Psychotherapy. In Manic
over the usual treatment regime. Evidence of which Depressive Illness (eds F. Goodwin & K. Jamison), pp. 725-745.
approaches will be most effective is lacking, so Oxford: Oxford University Press.
HARROW, M., GOLDBERG, J., GROSSMAN, L., et al (1990) Outcome
different psychotherapies need to be compared. It in manic disorders. Archives of General Psychiatry, 47,665-671.
will be important to distinguish specific and non HOOLEY, J., Ricwraits, J., WEIwrRAua, S., et a! (1987) Psycho
specific benefits (Lam, 1991). Short-term structured, pathology and marital distress: the positive side of positive
‘¿manualised'
therapiesadapted for use with BD, such symptoms. Journal of Abnormal Psychology, 96, 27-33.
JAMISON, K. & Axisit@, H. (1983) Medication compliance in patients
as cognitive therapy, IFI, BET and interpersonal with bipolar disorders. Psychiatric Clinics of North America, 6,
therapy, should be investigated initially. Follow-up 175—192.
in these studies must extend beyond the acute phase JAMISON, K., GERN1tR, R. & GooDwiN, F. (1979) Patient and
to assess any longer term benefits and, as drop-out physician attitudes towards lithium: relationshipto compliance.
Archives of General Psychiatry, 36, 866-869.
rates are likely to be high, recruitment will have to Joyca, P. (1992) Prediction of treatment response. In Handbook
be active and extensive to ensure that the sample is ofAffectiveDssorders(ed. E. S. Paykel), pp 453-464. London:
sufficiently large to reliably assess outcome. Churchill Livingstone.
 588 SCO1T

KAHN, D. (1990) The psychotherapy of mania. Psychiatric Clinics SHAKIR, S. VoLKMAN, F., BACON, S., et al (1979) Group psycho
of North America, 13, 229-240. therapy as an adjunct to lithium maintenance. American Journal
KRIPKE, D. & ROBINSON, D. (1985) Ten years with a lithium group. of Psychiatry,136, 455—456.
McLean Hospital Journal, 10. 1—11. SPENCER, J., GLICIC, I., HA@s, G., et a! (1988) A randomized
LM,i, D. (1991) Psychosocial family interventions in schizophrenia: control
trial
of in-patient
familyintervention.
III:
effects
at
a reviewof empirical studies. PsychologicalM&icine, 21,423-441. 6-month and 18-month follow-ups. American Journal of
MIKLOwIrz, D., GOLDSTEIN, M., NUEcHTERLEIN, K., et al (1988) Psychiatry, 145, 1115—1121.
Family factors and the course of bipolar affective disorder. TARGUM, S., DIBBLE, E., DAVENPORT, Y., et al (1981) The
Archivesof GeneralPsychiatry,45, 225-231. family attitude questionnaire. Patients' and spouses' views
Paar, M. & HARVEY, N. (1991) Lithium maintenance: 1. A of bipolar illness. Archives of General Psychiatry, 38,
standard education programme for patients. British Journal of 562—568.
Psychiatry, 158, 197-200. Vow.LAJt, F., SHAKIR,S., BACON, S., eta! (1981) Group therapy
Pmi, R. & P01mR, W. (1990)NIMH workshop report on treatment in the management
of manic-depressive
illness.American
of bipolar disorder. Psychopharmacology Bulletin, 26,409-427. Journal of Psychotherapy,42, 263—2671.
RUSH, A. (1988) Cognitive approaches to adherence. In Review of WULsIN,L., BACHOP,M. & Hosn@w@,D. (1988) Group therapy in
Psychiatry (vol. 8) (ads A. Frances & R. Hales), pp. 627-642. manic-depressive illness. American Journal of Psychotherapy,
Washington: American Psychiatric Association. 42, 263—271.

Professor Jan Scott, MRCPsych,

University Department of Psychiatry, Royal Victoria Infirmary, Newcastle
upon Tyne NE1 4LP

(First received 22 November 1994, final revision 6 March 1995, accepted 18 May 1995)

View publication stats