664794

research-article2016
ANP0010.1177/0004867416664794ANZJP ArticlesSportiche et al.

Research

Australian & New Zealand Journal of Psychiatry

Clinical factors associated with lithium 17

DOI: 10.1177/0004867416664794

response in bipolar disorders The Royal Australian and

New Zealand College of Psychiatrists 2016
Reprints and permissions:
sagepub.co.uk/journalsPermissions.nav
anp.sagepub.com

Sarah Sportiche1,2,3, Pierre Alexis Geoffroy1,2,3,4,5,

Clara Brichant-Petitjean1,2,3, Sebastien Gard3,6,
Jean-Pierre Khan3,7,8, Jean-Michel Azorin3,9,10, Chantal Henry3,11,12,
Marion Leboyer3,11, Bruno Etain1,2,3,4, Jan Scott13,14 and Frank Bellivier1,2,3,4

Abstract
Background: Bipolar disorder is a common chronic illness characterized by high levels of morbidity and all-cause
mortality. Lithium is one of the gold standard mood stabilizer treatments, but the identification of good, partial and non-
responders in clinical settings is inconsistent.
Methods: We used an established rating scale (the Alda scale) to classify the degree of lithium response (good response,
partial response, non-response) in a large, multicentre clinically representative sample of well-characterized cases of
bipolar disorders I and II. Next, we examined previously reported clinical predictors of response to determine which
factors significantly differentiated between the three response groups.
Results: Of 754 cases, 300 received lithium, for at least 6months, as a treatment for bipolar disorder (40%). Of these
cases, 17% were classified as good response, 52% as partial response and 31% as non-response. Lifetime history of mixed
episodes (p=0.017) and alcohol use disorders (p=0.015) both occurred in >20% of partial response and non-response
groups but <10% of good response cases. Family history of bipolar disorder I was of borderline statistical significance,
being more frequent in the good response group (38%) compared with the non-response group (18%). There was a
trend (p=0.06) for bipolar disorder II to be associated with non-response.
Conclusions: Only three factors previously identified as predictors of lithium response significantly differentiated the
response groups identified in our sample. Interestingly, these factors have all been found to co-occur more often than
expected by chance, and it can be hypothesized that they may represent a shared underlying factor or dimension. Further
prospective studies of predictors and the performance of the Alda scale are recommended.

Keywords
Bipolar disorder, lithium response, mixed state, alcohol misuse, clinical predictors

1Inserm, U1144, Paris, France

2AP-HP, GH Saint-Louis Lariboisire F. Widal, Ple de Psychiatrie et de Mdecine Addictologique, Paris, France
3Fondation FondaMental, Crteil, France
4Universit Paris Diderot and Sorbonne Paris Cit, UMR-S 1144, Paris, France
5Universit Paris Descartes, UMR-S 1144, Paris, France
6Hpital Charles Perrens, Centre Expert Bipolaire, Ple de Psychiatrie Gnrale Universitaire, Bordeaux, France
7Service de Psychiatrie et Psychologie Clinique, CHU de Nancy, Centre Psychothrapique de Nancy-Laxou, Nancy, France
8Universit de Lorraine, Nancy, France
9Ple de psychiatrie, Hpital Sainte Marguerite, Assistance Publique Hpitaux de Marseille, Marseille, France
10EA 3279-Self-perceived Health Assessment Research Unit, School of Medicine, Timone University, Marseille, France
11Inserm, U955, Equipe de psychiatrie gntique et Universit Paris Est, Facult de mdecine et AP-HP, Ple de psychiatrie, Hpitaux Universitaires

Henri Mondor, Crteil, France

12Institut Pasteur, Unit Perception et Mmoire, Paris, France
13Institute of Neuroscience, Newcastle University, Newcastle upon Tyne, UK
14Centre for Affective Disorders, Institute of Psychiatry, Psychology & Neuroscience (IoPPN), London, UK

Corresponding author:
Sarah Sportiche, AP-HP, GH Saint-Louis Lariboisire F. Widal, Ple de Psychiatrie et de Mdecine Addictologique, 75475 Paris, France.
Email: sarah.sportiche@aphp.fr

Australian & New Zealand Journal of Psychiatry

Downloaded from anp.sagepub.com at CORNELL UNIV WEILL MED COLG on August 26, 2016
2 ANZJP Articles
Introduction
Bipolar disorder (BD) is a common chronic illness charac-
terized by a high risk of mood recurrences, with 80% of
patients experiencing a relapse within 2years of an index
mood episode (Goodwin and Jamison, 2007; National
Institute for Health and Care Excellence [NICE], 2014).
The disorder is also characterized by high levels of physical
comorbidity, premature death and a high risk of suicide,
with a 6- to 10-fold increased risk compared to the general
population (Hayes etal., 2015). Unsurprisingly, BD is
ranked as one of the 10 most burdensome disorders world-
wide (Collins etal., 2011; Lopez and Murray, 1998).
Despite these levels of morbidity and all-cause mortality,
several longitudinal observational studies have shown that
prophylactic treatment, especially with lithium, can reduce
relapse, suicide and premature death rates and improve
functional outcomes compared to no treatment or sub-opti-
mal treatment (Mller-Oerlinghausen etal., 1994).
Furthermore, in clinical settings, lithium remains a gold
standard treatment with a significantly greater reduction in
recurrence rates, regardless of the polarity, relative to pla-
cebo or to other anticonvulsant mood stabilizers
(BALANCE Investigators and Collaborators etal., 2010;
Geddes etal., 2004; Severus etal., 2014). Lithium is the
only mood stabilizer that has been shown empirically to be
associated with a reduction in the risk of suicide in patients
with BD (Cipriani etal., 2005; Goodwin etal., 2003).
Although lithium appears to be a highly efficacious
treatment for BD in research settings, response rates to lith-
ium prophylaxis in clinical practice are highly variable
between individuals. Rybakowski etal. (2001) reported
that about 30% of individuals treated with lithium are clas-
sified as excellent responders (e.g. demonstrating a total
absence of affective episodes over a 10-year period of lith-
ium prophylaxis), while Grof etal. (2002) noted that the
response rate may be increased further in first-degree rela-
tives of lithium responders. However, the corollary is that,
in day-to-day clinical practice, about 30% of cases are par-
tial responders (prescription of lithium is associated with
some improvement) and up to 40% of BD cases are lithium
non-responders (i.e. the introduction of lithium does not
modify the course of the disease) (Baldessarini and Tondo,
2000; Garnham etal., 2007). Given these findings, there
have been some efforts to identify the personal and clinical
characteristics of individuals who are more rather than less
likely to respond to lithium prophylaxis.
Studies that attempt to identify clinical factors associ-
ated with good outcome following the introduction of lith-
ium prophylaxis have produced conflicting results. Two of
the most quoted predictors of Good Response (GR) are the
history of prophylactic response to lithium in first-degree
relatives and the episodic course of the disorder with remis-
sion between episodes (Grof etal., 2002). Other clinical
predictors have also been proposed such as bipolar subtype,
Australian & New Zealand Journal of Downloaded
Psychiatryfrom anp.sagepub.com at CORNELL UNIV WEILL MED COLG on August 26, 2016
number of mood episodes prior to the introduction of lith-
ium, age at onset, episode index polarity, rapid cycling,
comorbidities (personality, substance use or anxiety disor-
ders) and the presence of atypical features of depression.
However, in an extensive review of 43 studies and 42 clini-
cal variables, Kleindienst etal. (2005) were only able to
identify five potential predictors of clinical response to
lithium. The two predictors of GR to lithium treatment
were episodic illness pattern characterized as mania-
depression-interval (or MDI) and older age of onset of
BD; the three predictors of poor response were higher num-
ber of pre-lithium hospitalizations, an episodic pattern of
depression-mania-interval (or DMI) and a continuous
cycling pattern. The authors noted that all these findings
should be interpreted with caution, as there were inconsist-
encies in how lithium response was defined in the different
studies. Moreover, subsequent studies have not replicated
these findings. For example, Garnham etal. (2007) reported
that lithium responders were more likely to have an early
rather than later age of onset, to have an episodic illness
pattern (but did not report any differences between MDI or
DMI) and to meet diagnostic criteria for BD I. Pfennig
etal. (2010) found that the probability of recurrence in lith-
ium-treated patients was negatively associated with the
presence of mood-incongruent psychotic symptoms, inter-
episode residual symptoms and rapid cycling. Kessing
etal. (2011) found that excellent lithium responders were
characterized by few earlier psychiatric hospitalizations, a
manic index polarity of BD and lower levels of somatic
comorbidity.
Overall, there is some agreement about the proportion of
cases who are likely to show a GR (about 30%), but there is
uncertainty about the putative predictors of this desired out-
come. This issue has been compounded by two factors: the
heterogeneity in the definition or measurement of lithium
response and the lack of clinical representativeness and/or
small size of study samples. These issues have been par-
tially resolved by the introduction of the Retrospective
Criteria of Long-Term Treatment Response in Research
Subjects with Bipolar Disorder (also referred to as the Alda
scale) (Grof etal., 2002). The Alda scale has been used in
several studies by Grof etal. (2002), Duffy etal. (2002) and
Garnham etal. (2007) and has recently been shown to have
moderate but substantial inter-rater agreement and reliabil-
ity in a study of over 1000 cases recruited from 29 centres
participating in the Consortium on Lithium Genetics
(ConLiGen) network (Manchia etal., 2013). We therefore
employed this assessment measure to a representative clini-
cal sample of adult cases of BD I and II who had been pre-
scribed lithium by clinicians at one of four clinical services
located across France. In this study, we examined the demo-
graphic and clinical characteristics associated with member-
ship of one of three previously defined lithium response
categories, namely, GR, Partial Response (PR) and Non-
Response (NR) (Grof etal., 2002; Manchia etal., 2013).
Sportiche et al. 3
Methods
Ethical approval was granted by an institutional review
board for a programme of research on BD, and all partici-
pants gave written informed consent for their de-identified
data to be used in the research programme, which included
a range of projects including database studies.
Sample
The original research sample comprised participants who
were recruited between 1995 and 2008 from four French
university affiliated departments of psychiatry (Paris XII,
Bordeaux, Marseille and Nancy). The inclusion criteria
were that the person (a) was aged 18years; (b) had a mood
disorder that met Diagnostic and Statistical Manual of
Mental DisordersFourth Edition (DSM-IV) criteria for BD
I or BD II (American Psychiatric Association, 1994); (c)
currently met criteria for euthymia, which was operational-
ized as having scores <5 on both the Montgomerysberg
Depression Rating Scale (Montgomery and Asberg, 1979)
and the Bech Mania Rating Scale (Bech etal., 1978); and
(d) was willing and able to give written informed consent.
Assessments
Cases meeting the above inclusion criteria were assessed
by psychiatrists trained in the use of the French version of
the Diagnostic Interview for Genetic Studies (DIGS),
which is the equivalent of the Structured Clinical Interview
for DSM-IV, lifetime version (SCID-1) (Nurnberger etal.,
1994). The schedule (supplemented by additional questions
as appropriate) was used to record key socio-demographic,
family, clinical and treatment history variables in a research
database.
For the purposes of the current study, we identified BD
cases who were prescribed lithium as a prophylactic treat-
ment for at least 6months. The records of these cases were
then rated using the Alda scale (Grof etal., 2002) by a psy-
chiatrist (C.B.-P.) who had participated in the ConLiGen
reliability study (Manchia etal., 2013). The Alda scale was
specifically developed to allow retrospective evaluation of
prophylactic treatment response in naturalistic conditions;
it comprises two subscales (A and B). The A scale score
(range: 010) derives from items that quantify the degree of
improvement due to lithium treatment (e.g. reduction in the
frequency of recurrences, reduction in residual symptoms,
etc.). The B scale includes five criteria (each rated 02,
with a higher score being more negative) that represent fac-
tors that undermine the likelihood that the observed
improvement is really due to the introduction of lithium.
The five potential confounders incorporated in the B scale
are as follows: the number of episodes before receiving
lithium (B1), the frequency of episodes before receiving
lithium (B2), the duration of lithium treatment (B3), the
Downloaded from anp.sagepub.com at CORNELL UNIV WEILL MED COLG on August 26, 2016
compliance during periods of stability (B4) and the use of
additional medication during the period of stability (B5).
The total score (TS) on the Alda scale is obtained by sub-
tracting the B score from the A score; any negative score
(i.e. the B scale score exceeds the A scale score) is recorded
as 0. For the purposes of this study, we used the three
reported categories of response: NR (TS<2), PR (TS=2
6) and GR (TS7).
Many of the variables recorded in clinical case notes (e.g.
pre- and post-lithium recurrence rates) are used to rate the A
and B scale scores for the Alda scale, so the between-group
comparisons we report here focus on those characteristics
that were not already utilized to make the Alda rating, but
that have been identified as potentially important in predict-
ing likelihood of response to lithium in previously published
studies. These included age at onset of BD and polarity of
first episode, duration of illness prior to lithium, subtype of
BD, lifetime history of mixed episodes, psychotic symp-
toms, rapid cycling, suicide attempts, anxiety disorders,
alcohol or substance use disorders (AUD and SUD), sea-
sonal illness pattern and family history of BD I or II.
Statistical analysis
Mean and standard deviations (SDs) and frequency data
including number (N) and percentages (%) were calculated
for each response category. These data take into account any
missing information (so each comparison gives the number
of cases with available data). Non-parametric descriptive
statistical tests (2 and MannWhitney U tests) were used to
compare the GR, PR and NR groups. All analyses were per-
formed using the SPSS software package, version 20, and as
all variables for the comparison were selected a priori, we
set statistical significance at p<0.05 for all tests.
Results
Of the 754 cases in the research database, 300 had been
prescribed lithium for at least 6months (40%).
The study sample comprised 121 (40%) males and 179
(60%) females. Their mean age was 45.4 years (SD
13.2years), and the mean age at onset of BD was 25.8years
(SD 10.4years). In all, 240 cases (80%) met criteria for BD
I, while 60 cases (20%) met criteria for the BD II subtype.
The cases had experienced a mean of 8.5 BD episodes (SD
8.2), of which 4.9 (SD 5.1) were depressive, 3.2 (SD 4.3)
were manic and 0.2 (SD 0.4) were mixed episodes.
Individual had been ill for about 1012years prior to the
introduction of lithium.
As shown in Table 1, 17% of the sample was classified
as GR, 52% as PR and 31% as NR. Neither current age nor
gender was associated with response category. Family his-
tory of BD I was reported in 252 cases and was of border-
line statistical significance, being more frequent in the GR
group (17 out of 45 patients: 38%) compared with the PR
Australian & New Zealand Journal of Psychiatry
4 ANZJP Articles

Table 1. Clinical characteristics of groups categorized by Good Response (GR), Partial Response (PR) and Non-Response (NR) to
lithium (see text for details).

GR (n=50) PR (n=156) NR (n=94)

Variables N a (300) Mean (SD) or N (%)b U/2 Significance (p)

Females 300 31 (62%) 95 (61%) 53 (56%) 0.63 0.73

Current age in years (mean, SD) 300 47.6 (12.7) 44.6 (13.1) 45.4 (13.7) 2.53 0.81

Age at onset in years (Mean, SD) 297 27.4 (11.9) 25.4 (10.1) 25.5 (10.2) 0.57 0.75

Duration of illness before lithium 234 10.0 (9.0) 12.0 (10.3) 13.0 (11.0) 2.59 0.27
prescribed in years (mean, SD)

Bipolar I subtype 298 37 (74%) 128 (27%) 73 (79%) 1.89 0.39

Manic polarity at onset 297 24 (48%) 55 (36%) 34 (36%) 2.54 0.28

Seasonal patternc 268 4 (9%) 27 (19%) 22 (26%) 5.37 0.068

Psychotic symptomsc 297 50 (17%) 155 (52%) 92 (31%) 0.23 0.89

Mixed episodesc 269 3 (7%) 30 (21%) 23 (28%) 8.08 0.017

Rapid cyclingc 295 2 (4%) 19 (12%) 14 (15.2) 3.99 0.14

Attempted suicidec 293 24 (48%) 69 (45%) 42 (47%) 0.22 0.90

Anxiety disordersc 225 32 (45%) 51 (44%) 19 (19%) 0.02 0.99

Alcohol use disordersc 294 4 (8%) 26 (17%) 25 (27%) 8.36 0.015

Substance use disordersc 294 8 (16%) 33 (22%) 29 (32%) 5.07 0.08

Family history of Bipolar I disorder 252 17 (38%) 40 (31%) 14 (18%) 5.98 0.05

Family history of Bipolar II disorder 266 2 (4%) 9 (7%) 13 (15%) 5.73 0.06

SD: standard deviation.

aN included in analysis,
bPercentages are rounded to whole numbers.
cLifetime history.

(40 out of 131 patients: 31%) and NR groups (14 out of 76
patients: 18%); there was a non-significant trend (p=0.06)
for BD II to be associated with NR. Only two other factors
differed significantly between the three groups: lifetime
history of mixed episodes (p=0.017) and AUD (p=0.015);
these problems both occurred in >20% of PR and NR
groups but <10% of GR cases.
There were no statistically significant differences
between group with regard to mean age at the onset of BD,
or duration of illness prior to lithium, the characteristics of
the clinical presentation (bipolar subtype, onset polarity,
history of seasonal illness pattern, rapid cycling, psychotic
symptoms, suicide attempts) or the lifetime history of other
psychiatric comorbidities (anxiety disorder or SUD).
Discussion
Our main objective was to identify clinical factors associ-
ated with different levels of lithium response in a large,
clinically representative sample of patients with BD. We
identified 300 well-characterized cases of BD I and II who
were prescribed lithium, for at least 6months, at four
French centres, making this one of the biggest independent
studies of this issue. Of the previously purported predictors
of lithium response, only three clinical factors differed sig-
nificantly between the three predefined response groups:
family history of BD I, lifetime history of mixed episodes
and lifetime history of an AUD. Furthermore, the findings
on family history of BD I should be treated with caution as
data with regard to this variable were missing in 17% of the
sample and the statistical significance was borderline
(p<0.05) and not corrected for multiple testing.
The finding that lifetime history of AUD and lifetime
history of mixed episodes were both more common in
groups associated with lower response or NR to lithium
concurs with previous studies. For example, the response
rate to lithium prophylaxis has been shown to be poorer in
patients with BD and AUD (Frye and Salloum, 2006; Tohen
etal., 1990). The mechanism of this lithium resistance is
unexplained, but there are a number of hypotheses. First,

Australian & New Zealand Journal of Downloaded

Psychiatryfrom anp.sagepub.com at CORNELL UNIV WEILL MED COLG on August 26, 2016
Sportiche et al. 5
inadequate adherence may impact the response to lithium,
and it is well known that individuals with AUD and SUD
often show lower adherence rates than other patient popula-
tions (Aagaard and Vestergaard, 1990; Teter etal., 2011).
However, other evidence indicates that lithium adherence
in BD cases with comorbid AUD does not necessarily
improve response. For example, patients with BD and
comorbid AUD often have a younger age at onset for BD
(Feinman and Dunner, 1996; Winokur etal., 1998), a higher
risk of self-harm, more severe symptoms and more frequent
episodes (OConnell etal., 1991; Tohen etal., 1990) and
more rapid cycling and mixed episodes (Carvalho etal.,
2014); all of these characteristics may be associated with
poorer response to lithium. Furthermore, in animal studies,
a recent study has demonstrated that alcohol and lithium
have opposing effects on behavioural circadian rhythms
(Nascimento etal., 2015).
In line with our findings, previous studies have reported
poorer lithium response in BD cases with a lifetime history
of mixed episodes (Fountoulakis etal., 2012). Relatively
few studies have investigated the long-term medication in
patients with mixed episodes, but several suggest that lith-
ium may be less effective for these patients than some of
the other mood stabilizers such as valproate (Fountoulakis
etal., 2012; Freeman etal., 1992; Montgomery etal., 2000;
Yatham etal., 2013).
Finally, it is noteworthy that family history of BD I,
AUD and mixed states often co-occur, and thus, the three
predictors we have identified may represent different
aspects of some underlying genetic or pathophysiological
mechanism. A family history of BD has been noted as a
predictor of response to lithium in several studies (Grof,
2010; Maj etal., 1985; Mendlewicz etal., 1973), although
it is not always clear whether the important element is sub-
type of BD (I or II), family history of lithium response (i.e.
pharmaco-genetic aspects) or the pattern of illness in famil-
ial cases (such as offspring presenting with a more classic
Kraepelian form of manic depression). However, studies
have also noted that alcohol misuse, especially in males
with BD, is often more frequent in those with family his-
tory of BD (Frye etal., 2003), and AUD appears to increase
the risk of syndromal or sub-syndromal mixed states
(Paykel etal., 2006).
Several limitations of this study must be acknowledged.
The major weakness is the retrospective nature of the
assessment of lithium response. Of course, to a certain
extent, this is the only way to estimate the overall response
to lithium in a large clinical sample. However, a prospec-
tive study with contemporaneous recording of clinical data
pre- and post-lithium prescribing would allow more
detailed analysis of other potentially important predictors
and reduce any false-negative results that could have
occurred in our study because of random missing data (that
may have reduced the statistical power of some of our anal-
yses or between-group differences). This would be
Downloaded from anp.sagepub.com at CORNELL UNIV WEILL MED COLG on August 26, 2016
especially helpful in assessing each of the B scale items and
would offer a more reliable measure of the impact of fluc-
tuating levels of mediation adherence and/or variations in
therapeutic levels of lithium to be monitored. In this study,
we note that only 4% of our cases (N=11) had a high score
on the compliance item (B4 item score of 2). However,
this does not mean that all other cases had good compliance
throughout. Likewise, missing information from case notes
can hamper accurate assessment of B scale items, such as
detailed recordings of nature and severity of comorbidities
and their treatments.
On a positive note, this study identifies several new lines
for future research on predictors to lithium response, such as
clarifying the nature of familial response (and the best ele-
ment to choose to predict benefit from lithium) and the inter-
relationships between the most robust predictors we found. In
addition, it is clear that while the A and B subscales of the
Alda scale are admirably pragmatic and make clinical sense,
our experience of applying the scale to large clinical samples
leads us to believe that there may be some benefit in review-
ing the performance of the scale in more detail to determine
whether further refinements could improve its ability to iden-
tify valid clinical phenotypes of lithium response for use in
genetic and other studies (Schulze etal., 2010).
Conclusion
In our large French sample of individuals with BD I and II
who received lithium for at least 6months and who were
characterized for lithium response using the Alda scale, we
observed three factors associated with lithium response:
family history of BD I, lifetime history of mixed episodes
and lifetime history of an AUD. These three characteristics,
previously identified as predictors of lithium response in
the literature, have also been found to co-occur more often
than expected by chance and may represent a shared under-
lying factor or dimension. Future prospective studies
should test the positive and negative predictive value of this
cluster of characteristics on lithium prophylactic response.
Another finding, albeit weaker, is still noteworthy, namely
that family history of BD I appeared to be associated with
improving response, but family history of BD II showed a
trend in the opposite direction suggesting the need to
explore these separately rather than as a single construct.
Acknowledgements
S.S., F.B. and P.A.G. were involved in the design and planning of
the study. S.S., F.B., P.A.G. and J.S. identified the hypotheses for
the current study. C.B.-P. undertook the lithium response assess-
ments using the Alda Scale. S.S. carried out the literature review
and drafted the main sections of the manuscript. S.S., P.A.G. and
J.S. undertook the statistical analyses. S.G., J.P.-K., J.-M.A., C.H.,
B.E., M.L. and F.B. were responsible for the recruitment. F.B.,
B.E. and M.L. were, respectively, principal investigator and scien-
tific coordinators of the research project. F.B., P.G. and J.S.
Australian & New Zealand Journal of Psychiatry
6 ANZJP Articles
assisted S.S. in writing the preliminary draft. All authors read and
approved the final manuscript.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest
with respect to the research, authorship, and/or publication of this
article: S.S., C.B.-P., J.P.-K. declare that there is no conflict of inter-
est. P.A.G. has received travel awards or financial compensation
from AstraZeneca, Lundbeck, Menarini France and Otsuka. S.G.
has received honoraria and financial compensation as independent
symposium speakers from AstraZeneca, Bristol Myerrs Squib and
Otsuka. J.-M.A. has received research support and has acted as con-
sultant and/or serves on a speakers bureau for Bristol-Myers
Squibb, Eli Lilly, Lundbeck, Otsuka, Takeda, Novartis, Pfizer,
AstraZeneca, Servier and Sanofi-Aventis. C.H. has received hono-
raria and financial compensation as independent symposium speak-
ers from AstraZeneca, Bristol-Myers Squibb, Otsuka. M.L. reports
personal fees from Servier outside the submitted work. B.E. has
received honoraria and financial compensation as independent
symposium speakers from Lundbeck, AstraZeneca, Bristol-Myers
Squibb, Otsuka and Servier. F.B. has received honoraria and finan-
cial compensation as independent symposium speakers from
Sanofi-Aventis, Lundbeck, AstraZeneca, Eli Lilly, Bristol-Myers
Squibb and Servier and has received peer review research funding
from French Ministry of research, Assistance Publique Hpitaux
de Paris, the National Institute for Research (INSERM) and the
NARSAD. J.S. is a visiting professor at Diderot University. J.S. has
received grant funding from the Stanley Foundation (for work on
lithium and medication adherence), from the Medical Research
Council UK (including for projects on actigraphy and bipolar disor-
ders) and from the Research for Patient Benefit programme UK
(PB-PG-0609-16166: Early identification and intervention in young
people at risk of mood disorders). The FondaMental foundation has
received funding from Sanofi-Aventis (sponsorship of the Psy COH
BP cohort project), Roche Laboratory (sponsorship of the Autism
InfoR cohort), Otsuka and Lundbeck laboratories (FondaMental
Conferences), Eli Lilly & Co. (FondaMental PhD grant) and Servier
(FondaMental Post Doctoral grant).
Funding
The author(s) disclosed receipt of the following financial support
for the research, authorship, and/or publication of this article: This
research was funded by INSERM and Assistance Publique des
Hpitaux de Paris (Research Project number C0829 and P111002,
Ethics Approval number IDRCB2008-AO1465-50).
References
Aagaard J and Vestergaard P (1990) Predictors of outcome in prophylac-
tic lithium treatment: A 2-year prospective study. Journal of Affective
Disorders 18: 259266.
American Psychiatric Association (1994) Diagnostic and Statistical
Manuel of Mental Disorder, 4th Edition. Arlington, VA: American
Psychiatric Publishing.
BALANCE Investigators and Collaborators, Geddes JR, Goodwin GM,
etal. (2010) Lithium plus valproate combination therapy versus mon-
otherapy for relapse prevention in bipolar I disorder (BALANCE): A
randomised open-label trial. The Lancet 375(9712): 385395.
Baldessarini RJ and Tondo L (2000) Does lithium treatment still work?
Evidence of stable responses over three decades. Archives of General
Psychiatry 57: 187190.
Australian & New Zealand Journal of Downloaded
Psychiatryfrom anp.sagepub.com at CORNELL UNIV WEILL MED COLG on August 26, 2016
Bech P, Rafaelsen OJ, Kramp P, etal. (1978) The Mania Rating Scale:
Scale construction and inter-observer agreement. Neuropharmacology
17: 430431.
Carvalho AF, Dimellis D, Gonda X, etal. (2014) Rapid cycling in bipolar
disorder: A systematic review. The Journal of Clinical Psychiatry 75:
578586.
Cipriani A, Pretty H, Hawton K, etal. (2005) Lithium in the prevention of
suicidal behavior and all-cause mortality in patients with mood disor-
ders: A systematic review of randomized trials. The American Journal
of Psychiatry 162: 18051819.
Collins PY, Patel V, Joestl SS, etal. (2011) Grand challenges in global
mental health. Nature 475: 2730.
Duffy A, Alda M, Kutcher S, etal. (2002) A prospective study of the off-
spring of bipolar parents responsive and nonresponsive to lithium
treatment. The Journal of Clinical Psychiatry 63: 11711178.
Feinman JA and Dunner DL (1996) The effect of alcohol and substance
abuse on the course of bipolar affective disorder. Journal of Affective
Disorders 37: 4349.
Fountoulakis KN, Kontis D, Gonda X, etal. (2012) Treatment of mixed
bipolar states. The International Journal of Neuropsychopharmacology
15: 10151026.
Freeman TW, Clothier JL, Pazzaglia P, etal. (1992) A double-blind com-
parison of valproate and lithium in the treatment of acute mania. The
American Journal of Psychiatry 149: 108111.
Frye MA, Altshuler LL, McElroy SL, etal. (2003) Gender differences in
prevalence, risk, and clinical correlates of alcoholism comorbidity in
bipolar disorder. The American Journal of Psychiatry 160: 883889.
Frye MA and Salloum IM (2006) Bipolar disorder and comorbid alcohol-
ism: Prevalence rate and treatment considerations. Bipolar Disorders
8: 677685.
Garnham J, Munro A, Slaney C, etal. (2007) Prophylactic treatment
response in bipolar disorder: Results of a naturalistic observation
study. Journal of Affective Disorders 104: 185190.
Geddes JR, Burgess S, Hawton K, etal. (2004) Long-term lithium therapy
for bipolar disorder: Systematic review and meta-analysis of rand-
omized controlled trials. The American Journal of Psychiatry 161:
217222.
Goodwin FK and Jamison KR (2007) Manic-Depressive Illness. New
York: Oxford University Press.
Goodwin FK, Fireman B, Simon GE, etal. (2003) Suicide risk in bipo-
lar disorder during treatment with lithium and divalproex. Journal of
American Medical Association 290: 14671473.
Grof P (2010) Sixty years of lithium responders. Neuropsychobiology 62:
816.
Grof P, Duffy A, Cavazzoni P, etal. (2002) Is response to prophylactic lith-
ium a familial trait? The Journal of Clinical Psychiatry 63: 942947.
Hayes JF, Miles J, Walters K, etal. (2015) A systematic review and meta-
analysis of premature mortality in bipolar affective disorder. Acta
Psychiatrica Scandinavica 131: 417425.
Kessing LV, Hellmund G and Andersen PK (2011) Predictors of excellent
response to lithium: Results from a nationwide register-based study.
International Clinical Psychopharmacology 26: 323328.
Kleindienst N, Engel R and Greil W (2005) Which clinical factors predict
response to prophylactic lithium? A systematic review for bipolar dis-
orders. Bipolar Disorders 7: 404417.
Lopez AD and Murray CC (1998) The global burden of disease, 1990
2020. Nature Medicine 4: 12411243.
Maj M, Arena F, Lovero N, etal. (1985) Factors associated with response
to lithium prophylaxis in DSM III major depression and bipolar disor-
der. Pharmacopsychiatry 18: 309313.
Manchia M, Adli M, Akula N, etal. (2013) Assessment of response to
lithium maintenance treatment in bipolar disorder: A consortium on
lithium genetics (ConLiGen) report. PLoS ONE 8: 65636.
Mendlewicz J, Fieve RR and Stallone F (1973) Relationship between the
effectiveness of lithium therapy and family history. The American
Journal of Psychiatry 130: 10111013.
Sportiche et al. 7
Montgomery SA and Asberg M (1979) A new depression scale designed
to be sensitive to change. The British Journal of Psychiatry 134:
382389.
Montgomery SA, Schatzberg AF, Guelfi JD, etal. (2000) Pharmacotherapy
of depression and mixed states in bipolar disorder. Journal of Affective
Disorders 59: S39S56.
Mller-Oerlinghausen BT, Wolf B, Ahrens B, etal. (1994) Mortality dur-
ing initial and during later lithium treatment: A collaborative study by
the international group for the study of lithium-treated patients. Acta
Psychiatrica Scandinavica 90: 295297.
Nascimento NF, Carlson KN, Amaral DN, etal. (2015) Alcohol and lith-
ium have opposing effects on the period and phase of the behavioral
free-running activity rhythm. Alcohol 49: 367376.
National Institute for Health and Care Excellence (NICE) (2014) The
assessment and management of bipolar disorder in adults, children
and young people in primary and secondary care, updated edition.
London. Available at: www.nice.org.uk
Nurnberger JI, Blehar MC, Kaufmann CA, etal. (1994) Diagnostic inter-
view for genetic studies: Rationale, unique features, and training.
NIMH genetics initiative. Archives of General Psychiatry 51: 849859.
OConnell RA, Mayo JA, Flatow L, etal. (1991) Outcome of bipolar
disorder on long-term treatment with lithium. The British Journal of
Psychiatry 159: 123129.
Paykel E, Abbott R, Morriss R, etal. (2006) Sub-syndromal and syndro-
mal symptoms in the longitudinal course of bipolar disorder. The
British Journal of Psychiatry 189: 118123.
Downloaded from anp.sagepub.com at CORNELL UNIV WEILL MED COLG on August 26, 2016
Pfennig A, Schlattmann P, Alda M, etal. (2010) Influence of atypical fea-
tures on the quality of prophylactic effectiveness of long-term lithium
treatment in bipolar disorders. Bipolar Disorders 12: 390396.
Rybakowski JK, Chlopocka-Wozniak M and Suwalska A (2001) The pro-
phylactic effect of long-term lithium administration in bipolar patients
entering treatment in the 1970s and 1980s. Bipolar Disorders 3: 6367.
Schulze TG, Alda M, Adli M, etal. (2010) The International Consortium
on Lithium Genetics (ConLiGen): An initiative by the NIMH and
IGSLI to study the genetic basis of response to lithium treatment.
Neuropsychobiology 62: 7278.
Severus E, Taylor MJ, Sauer C, etal. (2014) Lithium for prevention of
mood episodes in bipolar disorders: Systematic review and meta-
analysis. International Journal of Bipolar Disorders 2: 15.
Teter CJ, Falone AE, Bakaian AM, etal. (2011) Medication adherence
and attitudes in patients with bipolar disorder and current versus
past substance use disorder. Psychiatry Research 190: 253258.
Tohen M, Waternaux CM, Tsuang MT, etal. (1990) Four-year follow-
up of twenty-four first-episode manic patients. Journal of Affective
Disorders 19: 7986.
Winokur G, Turvey C, Akiskal H, etal. (1998) Alcoholism and drug abuse
in three groups Bipolar I, unipolars and their acquaintances. Journal
of Affective Disorders 50: 8189.
Yatham LN, Fountoulakis KN, Rahman Z, etal. (2013) Efficacy of aripipra-
zole versus placebo as adjuncts to lithium or valproate in relapse preven-
tion of manic or mixed episodes in bipolar I patients stratified by index
manic or mixed episode. Journal of Affective Disorders 147: 365372.
Australian & New Zealand Journal of Psychiatry