Eur Arch Psychiatry Clin Neurosci (2001) 251: Suppl.

2, II/72II/75

Steinkopff Verlag 2001

B. Mller-Oerlinghausen

Arguments for the specificity of the antisuicidal effect of lithium

Abstract Affective disorders are characterized by rst

a high recurrence risk, second a 3050 times increased
suicide risk and third a 2- to 3 times increased overall
mortality. In contrast to a populistic belief no scientic
evidence exists that antidepressant treatment, particularly long-term treatment, could reduce the the risk of
suicidal acts in depressive patients with a history of suicide attempts. Data, however, coming from international, systematic, retrospective analyses of well-documented long-term courses of illness in reliably diagnosed
patients, and from a large national, prospective longterm trial on the prophylactic efcacy of lithium versus
carbamazepine and amitriptyline has accumulated in the
last 10 15 years strongly supporting a (possibly specic)
antisuicidal effect of lithium. The large collaborative
IGSLI study (International Group for the Study of
Lithium-treated Patients) covering 5,616 patient years
clearly showed that adequate long-term lithium treatment significantly reduces and even normalizes the
excess mortality of patients with affective disorders. A
metaanalysis on 17,000 patients pooled from 28 studies
demonstrated that the rate of suicidal acts is 8.6 fold
higher in patients without lithium as compared to those
with regular lithium treatment.
A post-hoc analysis of a large multicenter, controlled
long-term trial found no suicidal acts in 146 patients
randomized to lithium compared to 9 suicidal acts in 139
patients randomized to carbamazepine.
Reanalysis of the data from the IGSLI study supports
the concept of the specicity of lithium, i.e., evidence
could be provided that lithium also reduces suicidal
behavior in patients who do not benet from the lithium
treatment in terms of episode reduction.

Prof. Dr. med. B. Mller-Oerlinghausen ()

Jebensstr. 3, 10623 Berlin, Germany
Tel.: +49-30-31001-361
Fax: +49-30-31001-366
E-Mail: bmoe@zedat.fu-berlin.de

Conclusion Lithium has to be considered as a rst line

mood stabilizer in affective disorders, particularly in
patients with a history of suicide attempts. Extreme caution is required when lithium is discontinued or a patient
is switched to another mood stabilizer, because such a
patient might have been protected against suicidal
impulses in spite of an incomplete response as to the
number and quality of depressive/manic episodes.

Key words Suicide lithium specificity carbamazepine affective disorders
Dedicated to Prof. Per Bech in appreciation and admiration for his continuous support and promotion of a rational and optimized pharmacotherapy in psychiatry.

Introduction
Affective disorders are characterized by 1) the high risk
of recurrences and 2) the 2- to 3-fold increased mortality
rate which is mainly caused by the 30 50 times higher
risk of suicide. For many years it was a silent, apparently
self-evident assumption that vigorous treatment of acute
depressive episodes and antidepressive long-term treatment would also reduce the suicide risk and the excess
mortality. However, in spite of the worldwide use of antidepressants, epidemiological studies hardly supported
such hope. The question, thus, arises: are there any pharmacological strategies to prevent suicidal acts and to
reduce the overall excess mortality? In fact, a number of
ndings from the past 15 years, coming from various
sources, various groups and various countries have accumulated sound evidence that appropriate lithium longterm medication, administered and monitored according to the state of the art, does decrease the suicide risk
and thus reduces or even normalizes the excess mortality in bipolar and schizoaffective as well as unipolar
patients.

II/73

Historical aspects of the antisuicidal effect of lithium

How did the detection of this fascinating antisuicidal
effect come about?
Based on observations made in our lithium clinic in
1985 we began a systematic follow-up of a group of
patients with a high suicide risk, i.e., those with at least
one suicide attempt in their history before onset of
lithium prophylaxis and with lithium medication maintained for at least one year. We observed a very marked
reduction of suicide attempts which intriguingly
occurred not only in responders towards lithium prophylaxis in terms of episode reduction but also in
non-responders. In addition, many more suicidal acts
were committed by patients having discontinued lithium
than by those under regular lithium treatment (Table 1).
At that time in the early 1980s only scattered information was available in the literature on a possible association of lithium prophylaxis and change of suicidality.
In addition, Barraclough (1972) had observed that suicide victims sufferinged from mental illness often had
received inadequate drug treatment. He estimated that,
in his sample, 20 of 100 suicides could have prevented by
appropriate lithium treatment. The best documented
information came from Hanus and Zapletlek (1984).
They reported that whereas 25 out of 99 patients with
affective disorders tried to commit suicide in a period of
5 years preceding the start of lithium treatment, only 4
suicide attempts and no suicides occurred within a 5years course of lithium treatment.
After the fall of the wall between West and East Germany we became aware of the ndings of Felber et al.
(1994) from Dresden, who had observed practically identical changes as we had in the former West Berlin, but
during the time of socialism he had not been allowed to
publish his ndings. The Germany ndings were also
supported by a publication from the U.K. (Coppen et al.
1991) referring to mostly unipolar patients (Table 2).
These preliminary ndings were the beginning of a
very fruitful international scientic collaboration over

Tab. 1 Main findings of the first Berlin study on the antisuicidal properties of lithium
(Causemann and Mller-Oerlinghausen 1988; Mller-Oerlinghausen et al. 1992)
Lithium treatment > 1 year N = 68
No. of suicide attempts before lithium
treatment
55 patients on regular lithium treatment
13 discontinuers of lithium

2.1
0.2
2 suicides
4 suicide attempts
4 suicides
7 suicide attempts

Mller-Oerlinghausen et al. (1992)

Felber and Kyber (1994)
Coppen et al. (1991)

On lithium

Off lithium

1 of 55
1 of 36
1 of 103

4 of 13
3 of 36
13 of 103*

* nontreatment group

the past 15 years. In 1986, Mogens Schou had started the

International Group for the Study of Lithium-treated
Patients (IGSLI) which included 6 specialized lithium
centers in different countries. The outcome criterion of
the rst IGSLI mortality study was the standardized mortality ratio (SMR), i.e., the death rate of a patient sample
is compared with the odds for dying in an exactly individually matched sample of the general population
(Mller-Oerlinghausen et al. 1992). The IGSLI group
found an SMR of 0.89, i.e., it did not differ from the mortality of the normal population. We could also show in
successive studies that the mortality increased after discontinuation of lithium to the expected values in
untreated patients with affective disorders (MllerOerlinghausen et al. 1992). Nilsson (1995), although she
did not observe a full normalization of the SMR, also
showed a signicant increase after discontinuation of
lithium up to the expected gure of around 3 (Table 3).
Tab. 3 Standardized mortality rate (SMR) of patients with affective disorders during
appropriate lithium treatment and after its discontinuation
SMR

Mller-Oerlinghausen et al. (1992)

Lenz et al. (1994)
Nilsson (1995)

During lithium
treatment

After
discontinuation

0.89
0.86
1.8***

2.54**
1.8*
3.1***

(SMR 1.0 indicates that mortality does not differ from that of the general population) significantly different from 1.0: *p < 0.05, **P < 0.01, ***p < 0.001

Reanalyzing the data from the IGSLI main study

allowed us to differentiate between the suicide mortality
and other causes of mortality. As Table 4 shows, the specic suicide-induced mortality in contrast to the cardiovascular mortality is not fully normalized, but it is
Tab. 4 Standardized mortality rate (SMR) of 5616 patient years (average lithium
treatment time 81 months) (Ahrens et al. 1995)

No. of suicide attempts/patient

Before lithium

During lithium

Responders n = 42
Non-responders n = 26

1.3
0.2
1.5
0.2

0.00**
0.23
0.13**

** = P < 0.01

Tab. 2 Completed suicides in patients on and off lithium

All deaths
Suicides
Cardiovasc deaths
Other

SMR

95 % conf. interv.

Expected

1.14
5.22
0.93
1.04

0.74 1.69
1.70 12.18
0.45 1.71
0.64 1.61

ca. 2 3
ca. 30 78
ca. 3 8

II/74

about one tenth of what one would expect in untreated

patients (Ahrens et al. 1995).
In the meantime, other groups have replicated and
supported these original observations, and it should be
emphasized that at present no study exists which would
have falsied the concept of a suicide preventive effect of
lithium long-term treatment (Schou 1998). Tondo et al.
(1997) reviewed the existing evidence and reported that
there is an 8.6-fold increased risk of suicidal acts in
patients off lithium as compared to those on lithium.
These data are extracted from 28 studies covering 17,000
patients. Further support comes from a recent and careful Swedish study by Kallner et al. (2000) which refers to
more than 6,000 patient years and which supports the
IGSLI analysis of 5,600 patient years, since according to
these authors the suicide- induced mortality is still
increased (SMR = 14.0), at least in bipolar patients, but is
doubled (SMR = 33) in patients off lithium. It also indicates that patients being treated in a specialized lithium
clinic have a lower SMR than those being taken care by
GPs, psychiatrists in private practices, etc.

Specificity of lithium?
In view of these rather robust and consistent empirical
ndings, the intriguing question arises whether the suicide- preventive effect of lithium should be considered a
specic effect and what could be the underlying mechanism?
For the sake of clarity we may subdivide the issue of
potential specicity into two questions:
1. Is this effect specic for lithium salts, i.e., is it not
shared by other drugs, such as other mood stabilizers
or antidepressants, etc.?
2. Is this effect strictly related to the episode-preventive
effect of lithium prophylaxis or might it act independently?
The question whether the antisuicidal effect is shared by
other psychotropic agents was addressed in the German
MAP study, a prospective, randomized, controlled trial
with a treatment duration of 2.5 years. 146 bipolar and
schizoaffective patients were randomized to lithium, 139
to carbamazepine. No suicidal act was observed in the
lithium group, 4 suicides and 5 suicide attempts occurred
in the carbamazepine group a statistically signicant
difference (Thies-Flechtner et al. 1996; Ghaemi and
Goodwin 1999).
Another argument could be taken from the study by
Modestin and Schwarzenbach (1992) who were running
a follow-up of 64 former psychiatric inpatients who had
committed suicide within one year after discharge and
compared them with a carefully matched control group
of patients who had not committed suicide. A signicantly higher proportion of the controls had been receiving various kinds of psychopharmacotherapy including
seven patients who had been treated with lithium. How-

ever, none of the 64 patients who committed suicide had

been receiving lithium at the time of his or her death the
only statistically signicant difference between the two
groups.
What could be the mechanism of this suicide-preventive effect? Our hypothesis from the very beginning has
always been that lithium differs from other mood stabilizers and also from most antidepressants by its very
marked serotonin-agonistic effects which are predominantly related to its pre-synaptic functions. It appears at
least an attractive speculation that this serotoninergic
action, possibly in connection with other effects, is
related to its very well-established antiaggressive effects
in animals as well as humans, but also to its antisuicidal
effects.
The second and decisive question is, whether the antisuicidal effect of lithium would also occur in patients not
responding optimally in terms of episode prevention.
This question again implies two different issues:
a) Are there neurobiological concepts and epidemiological data to support the idea of suicidal behavior as an
independent nosological entity?
b) Does lithium effectively reduce suicidal behavior also
in patients who do not benet from the lithium treatment in terms of episode reduction?
In fact, some concepts and ndings suggest that suicidal
behavior might be seen as a particular, possibly angerrelated form of affective dysregulation, also associated
with a disturbance of the 5HAT system, and, thus, as an
independent nosological syndrome.
Data from the recent large WHO study shows that the
prevalence of suicidal behavior is not fully related to the
existence of ICD psychiatric diagnoses but that it occurs
frequently in symptomatic individuals and in subjects
with subthreshold disorders. Many such people, according to WHO data from Germany, are characterized by
symptoms of overt or suppressed anger which van Praag
(1992) considers as one of the core constituents of the
stress syndrome, together with anxiety. Van Praag postulated that in certain types of depression characterized
by a 5HT-decit anxiety and aggression regulation is
primarily disturbed while mood lowering is a derivative
symptom. Consequently, he expected that certain drugs
such as l-tryptophan, the azapirones or lithium might
ameliorate anxiety and/or aggression regulation via regulation of the 5HT system to exert in addition an overall
therapeutic effect in depression.
As to the second aspect, namely whether the antisuicidal effect of lithium would also occur in patients not
responding optimally in terms of episode prevention,
part of the IGSLI data does support such a concept.
For this subanalysis we selected only patients with at
least one suicide attempt in the past before onset of
lithium medication (n = 176). We divided the sample
into three subgroups according to their response to
lithium long-term treatment in terms of reduction of
depressive in-patient episodes. Despite the clearly different overall efcacy of lithium prophylaxis a statistically

II/75
Tab. 5 Specificity of lithium. Reanalysis of IGSLI data from 176 high risk patients
(Ahrens and Mller-Oerlinghausen 2001)

Poor response (n=41)

Questionable response
(n=81)
Excellent response
(n=45)

No. of suicide
attempts per year
before lithium
treatment

No. of suicide
attempts per year
during lithium
treatment

0.33

0.10

p < 0.007

0.27

0.06

p < 0.000

0.26

0.02

p < 0.000

signicant reduction of suicide attempts occurred in all

three groups, even in the poor responders who did not
show a signicant decrease of the depressive in-patient
episodes (Table 5). In other words, in 50% of the clearcut non-responders no other further suicide attempt was
observed during lithium treatment.
The standardized suicide mortality in the poor
responders was 17.0 compared to an expected gure of
approx. 100. Certainly, these ndings can neither prove
the suicide-preventive effect of lithium nor its potential
specicity. However, the accumulated evidence strongly
supports such a hypothesis.
Which consequences should be drawn from our present state of knowledge? We feel that, rst, the potential
antisuicidal effect should be considered in setting up
modern guidelines for maintenance treatment in affective disorders (Mller-Oerlinghausen et al. 2001). Furthermore, doctors should be extremely cautious in discontinuing lithium because of alleged non-response.
Despite insufcient episode reduction the patient may
have been protected against suicidal impulses. Therefore, it could be wiser to add a second mood stabilizer
instead of switching the patient to a monotherapy with
e.g. carbamazepine.

References
Ahrens B, Mller-Oerlinghausen B (2001) Does lithium exert an independent antisuicidal effect? Pharmacopsychiatry 34: 132136

Ahrens B et al. (1995) Excess cardiovascular and suicide mortality of

affective disorders may be reduced by lithium prophylaxis. J Affect
Disord 33: 6775
Barraclough B (1972) Suicide prevention, recurrent affective disorder
and lithium. Br J Psychiatry 121: 391392
Causemann B, Mller-Oerlinghausen B (1988) Does lithium prevent
suicides and suicidal attempts? In: Birch NJ (Ed) Lithium: Inorganic
Pharmacology and Psychiatric Use. IRL Press Limited, Oxford, pp
2324
Coppen A et al. (1991) Does lithium reduce the mortality of recurrent
mood disorders? J Affect Disord 23: 17
Felber W, Kyber A (1994) Suizide and Parasuizide whrend und auerhalb einer Lithumprophylaxe. In: Mueller-Oerlinghausen B, Berghoefer A (Eds) Ziele und Ergebnisse der medikamentsen Prophylaxe affektiver Psychosen. Thieme, Stuttgart, pp 5359
Ghaemi SN, Goodwin FK (1999) Use of atypical antipsychotic agents in
bipolar and schizoaffective disorders: review of the empirical literature. J Clin Psychopharmacol 19: 354361
Hanus K, Zapletlek M (1984) Sebrevrazedn aktivita nemocnych
aktivnimi poruuchmi v prbehu lithioprophylaxe. Ceskoslovensk
Psychiatrie 80: 97100
Kallner G et al. (2000) Mortality in 497 patients with affective disorders
attending a lithium clinic or after having left it. Pharmacopsychiatry 33: 812
Lenz G et al. (1994) Mortalitt nach Ausscheiden aus der Lithiumambulanz. In: Mueller-Oerlinghausen B, Berghoefer A (Eds) Ziele und
Ergebnisse der medikamentsen Prophylaxe affektiver Psychosen.
Thieme, Stuttgart, pp 4952
Modestin J, Schwarzenbach F (1992) Effects of psychopharmacotherapy
on suicide risk in discharged psychiatric inpatients. Acta Psych
Scand 85: 173175
Mller-Oerlinghausen B et al. (1992) Suicides and parasuicides in a
high-risk patient group on and off lithium long-term medication. J
Affect Disord 25: 261270
Mller-Oerlinghausen B et al. (1992) The effect of long-term lithium
treatment on the mortality of patients with manic-depressive and
schizoaffective illness. Acta Psych Scand 86: 218222
Mller-Oerlinghausen B et al. (2001) Manic-depressive disorder. The
Lancet (in press)
Nilsson A (1995) Mortality in recurrent mood disorders during periods
on and off lithium: a complete population study in 362 patients.
Pharmacopsychiatry 28: 813
Schou M (1998) The effect of prophylactic lithium treatment on mortality and suicidal behaviour: a review for clinicians. J Affect Disord
50: 253259
Thies-Flechtner K et al. (1996) Effect of prophylactic treatment on suicide risk patients with major affective disorders. Pharmacopsychiatry 29: 103107
Tondo L et al. (1997) Effect of lithium maintenance on suicidal behaviour in major mood disorders. In: Stoff JM, Mann JJ (Eds) The Neurobiology of Suicide: From the Bench to the Clinic. Ann N Y Acad
Sci 836: 339351
van Praag HM (1992) Depression, aggression, and anxiety: a biological
hypothesis about their interrelation. Eur Neuropsychopharmacol 2:
393402