Review
Introduction
Methods
Results
Year/country | Author | Diagnoses | Medication/type of study | Number of patients | Duration | Results |
---|---|---|---|---|---|---|
1971/UK | Coppen et al. | Unipolar | Li | 65 | 2 years | 3 died (Plac) from causes not related to affective disorders |
Bipolar | Placebo | Li (n = 28) | ||||
Plac (n = 37) | ||||||
Randomly assigned | Unipolar Li (n = 11) Unipolar Plac (n = 17) Bipolar Li (n = 15) Bipolar Plac (n = 22) | |||||
1973/USA | Prient et al. (a) | Unipolar | Li | 122 | 2 years, observation period started after patient discharge | 1 suicide (Plac), 4 deaths due to other causes not related to affective disorder |
Bipolar | Imipramine | Li (n = 45) Imipramine (n = 38) Plac (n = 39) | ||||
Placebo | ||||||
Randomly assigned | Bipolar (n = 44) | |||||
Unipolar (n = 78) | ||||||
1973/USA | Prien et al. (b) | Bipolar | Li | 205 | 2 years | 1 suicide (Plac) |
Placebo | Li (n = 101) | 2 deaths due to other causes | ||||
Plac (n = 104) | ||||||
Randomly assigned | ||||||
1989/USA | Dorus et al. | Depressed vs. non-depressed alcoholics | Li compliant | 457 | 52 weeks | 1 death from all causes (Plac), no death (Li) |
(n = 82) | Depressed (n = 171 ≥ 108 study completers) | |||||
Placebo compliant (n = 82) | ||||||
Non-depressed (n = 286 ≥ 172 study completers) | ||||||
Double-blind, placebo-controlled trial | ||||||
1996/Germany | Greil et al. | Unipolar depression | Li | 81 | 2.5 years | 1 suicide (amitriptyline) 0 suicide (Li) |
Amitriptyline | Li (n = 40) | |||||
Prospective randomized multicenter trial | Amitriptyline (n = 41) | |||||
1997/Germany | Greil et al. (a) | Bipolar disorder | Li | 144 | 2.5 years | 0 suicide or SA (Li) |
CBZ | Li (n = 74) | 1 suicide, 1 SA (CBZ) | ||||
CBZ (n = 70) | ||||||
Prospective randomized multicenter trial | ||||||
1997/Germany | Greil et al. (b) | Schizoaffective disorder | Li | 90 | 2.5 years | 0 suicide or SA (Li) |
CBZ | Li (n = 43) | 0 suicide, 4 SA (CBZ) | ||||
CBZ (n = 47) | ||||||
Prospective randomized multicenter trial | ||||||
2000/UK | Coppen et al. | Unipolar | Li | 103 | Study population was recruited in 1977 | 10 patients died during the study, expected number of deaths was 18.31 and no deaths from suicide |
Bipolar | Placebo | Unipolar (n = 67) Bipolar (n = 30) Schizoaffective (n = 6) | ||||
Schizoaffective | Randomly assigned | |||||
2000/Germany | Bauer et al. | Refractory depression | AD + Li | 30 | 4.5 months | 1 suicide (Plac) |
AD + placebo | AD + Li (n = 14) | |||||
AD + Plac (n = 15) | ||||||
Randomized, double-blind, placebo-controlled trial | ||||||
2002/UK | Wilkinson et al. | Unipolar depression | AD + Li | 49 | 2 years | 4 deaths from unrelated causes (2 Li, 2 Plac) |
AD + placebo | AD + Li (n = 25) | |||||
AD + Plac (n = 24) | ||||||
Randomized, double-blind, controlled trial | Elderly patients (>65) | |||||
2003/Europe, Canada, USA | Calabrese et al. | Bipolar I | Li | 966 | 8–16 weeks open label, 18 months | 6 deaths: 4 suicides, 2 of the suicides during open-label study, 1 suicide after discontinuation from open-label study |
Placebo | Li (n = 121) | |||||
LTG | Plac (n = 121) | |||||
LTG (n = 221) | ||||||
Randomized, double-blind, placebo-controlled trial | ||||||
0 suicide (Li), 1 suicide (LTG) | ||||||
11 SA (10 open label, 1 in Plac) | ||||||
2005/USA | Findling et al. | Bipolar I, II | Li | 60 | Phase 1: up to 20 weeks | 0 suicide, 0 SA in both groups |
VLP | Li (n = 30) | |||||
VLP (n = 30) | Phase 2: 76 weeks | |||||
Randomized, double-blind, controlled trial | Pediatric (5–17 years) | |||||
2005/USA, Africa, Australia, Canada, Europe | Tohen et al. | Bipolar I | Li | 431 | Open label with Li + OLZ 6–12 weeks, 52 weeks double-blind phase monotherapy | 1 suicide (Li) during open-label phase, 2 died (Li) during double-blind phase (1 suicide, 1 accident) |
OLZ | Li (n = 214) | |||||
OLZ (n = 217) | ||||||
Randomized, double-blind, controlled trial | ||||||
2007/Netherlands | Kok et al. | Treatment resistant depression | Augmentation with Li | 29 | 6 weeks + 2 years follow-up | 2 deaths (Li), 3 deaths (phenelzine) |
Li (n = 15) | ||||||
Augmentation with phenelzine | Phenelzine (n = 14) | |||||
Elderly (>60 years) | ||||||
Open, randomized controlled trial | ||||||
2008/Germany | Lauterbach et al. | Depressive disorders | Li | 167 | 12 months | 0 suicide (Li) |
Placebo | Li (n = 84) | 3 suicides (Plac) | ||||
Plac (n = 83) | 7 SA (Li) | |||||
Randomized, double-blind, placebo-controlled trial | 7 SA (Plac) | |||||
2010/UK, France, USA, Italy | Geddes et al. | Bipolar I | Li | 330 | 24 months | 2 deaths from all causes (Li), 3 deaths from all causes (VLP), 1 death from all causes (Li + VLP) |
VLP | Li (n = 110) | |||||
Li + VLP | VLP (n = 110) | |||||
Li + VLP (n = 110) | ||||||
Randomized, open-label trial | ||||||
2010/Denmark, Sweden | Licht et al. | Bipolar I | Li | 155 | 52 weeks | 1 suicide (Li) |
LTG | Li (n = 78) | 1 SA (LTG) | ||||
LTG (n = 77) | ||||||
Open, randomized trial | ||||||
2011/USA | Oquendo et al. | Bipolar disorder | Li | 98 | 2.5 years | 0 suicide |
VLP | Li (n = 49) | 6 SA (Li), 8 SA (VLP) | ||||
Randomized, double-blind controlled trial | VLP (n = 49) | 45 suicidal events (16 Li, 18 VLP) | ||||
2011/Asia, Europe, Central and South America, USA | Weisler et al. | Bipolar I | Li | 2438 (open label) | Open label 4–24 weeks, double-blind up to 104 weeks | 3 deaths (Quet): 2 due to other causes, 1 suicide/accidental gunshot, similar overall incidence of suicidal behavior/ideation in the Quet (n = 3), Li (n = 3), and Plac (n = 8) groups |
Quetiapin | 1226 (double-blind) | |||||
Placebo | Li (n = 364) | |||||
Quet (n = 404) | ||||||
Randomized, placebo-controlled trial | Plac (n = 404) | |||||
2011/USA | Khan et al. | Depressive disorders | Cit + Li | 80 | 4 weeks | Subgroup of the patients assigned to citalopram and lithium had significantly higher Sheehan Suicidality Tracking Scale (S-STS) remission rates compared to patients assigned to citalopram and placebo |
Cit + Placebo | Cit + Li (n = 40) | |||||
Cit + Plac (n = 40) | ||||||
Randomized, double-blind trial |
Year/country | Author | Diagnoses | Medication/type of study | Number of patients | Duration | Results |
---|---|---|---|---|---|---|
1972/UK | Barraclough | Affective disorders | Retrospective, clinical survey | 100 suicide cases | >1/5 of these 100 suicides might have been prevented if lithium had been used | |
1977/USA | Fieve | Affective disorders | Li | 20 | 78 weeks | 0 suicide |
Retrospective cohort study | ||||||
1984/Czech Republic | Hanus and Zapletalek | Affective disorders | Li treatment compared to non-Li treatment/retrospective cohort study | 95 | 5 years | 4 SA during Li therapy |
20 % reduction | 25 SA during non-Li therapy | |||||
1991/UK | Coppen et al. | Affective disorders | Li | 103 | 11-year follow-up | 10 deaths from other causes, expected number 18.31; no suicides |
Retrospective cohort study | ||||||
1991/Denmark | Vestergaard and Aargaard | Affective disorders | Li | 133 | 5 years | No advantage |
Retrospective cohort study | ||||||
1992/Germany, Denmark, Canada, Austria | Müller-Oerlinghausen et al. (a) | Affective disorders | Li | 827 | Ø 81 months = 5600 patient-years | 44 deaths, total number of deaths was not different from what would be expected in a matched sample of the general population |
Retrospective cohort study | IGSLI cohort | |||||
1992/Germany | Müller-Oerlinghausen et al. (b) | Affective disorders | Comparison of Li treatment with non-Li treatment period | 68 | 8 years, at least 12 months of Li treatment | 2 suicides, 4 SA during Li treatment (in 55 patients) |
4 suicides, 7 SA when not Li-treated | ||||||
Retrospective cohort study | ||||||
1994/Germany | Felber and Kyber | Affective disorders | Comparison of Li treatment with non-Li treatment period | 36 | 14 years, 6.8 years with lithium, 7.2 years without lithium | 64 SA during non-Li treatment, |
7 SA during Li treatment, reduction SA 10:1, reduction suicides 3:1 | ||||||
Retrospective cohort study | ||||||
1995/Sweden | Nilsson | Affective disorders | Comparison of Li treatment with non-Li treatment period | 362 | At least 1 year of Li treatment | 129 deaths, risk to die was 1.7 times higher when not Li-treated and 4.8 times higher to commit suicide when not Li-treated |
Retrospective cohort study | ||||||
1996/Germany, Canada, Denmark, Austria, Sweden | Wolf et al. | Affective disorders | Li | 1056 | Average length of Li treatment 86.2 months | Mortality of lithium-treated patients did not differ from general population |
Retrospective cohort study | Total treatment period 90.982 months | |||||
IGSLI cohort | ||||||
1998/Italy | Bocchetta et al. | Affective disorders | Li | 100 | Ø 9.6 years observation period | 10 suicides—9 of them after discontinuation of Li therapy, suicide risk 24-fold higher when not Li-treated |
Retrospective cohort study | ||||||
1999/USA | Baldessarini et al. | Bipolar disorder | Li | 310 | Ø 6.36 years Li, of that group, 128 patients with Ø 3.7 years without lithium + 8.2 years observation period before lithium therapy | 7-fold lower rate of suicidal events during Li therapy compared to time before Li therapy rate of suicidal events 20-fold higher after discontinuation |
Retrospective cohort study | ||||||
2000/Sweden | Kallner et al. | Affective disorders | Li | 497 | 30 years | 80 % higher suicide rate when Li therapy was discontinued |
Retrospective cohort study | ||||||
2000/Denmark | Brodersen et al. | Affective disorders | Li | 133 | 2 years observation period + follow-up after 16 years | 40 deaths, including 11 suicides, mortality was twice that of general population (due to suicides), mortality from all other causes was similar; suicides occurred more often in those not compliant with therapy |
Retrospective cohort study | ||||||
2000/UK | Coppen et al. | Unipolar Bipolar Schizoaffective | Retrospective cohort study | 67 unipolar, 30 bipolar, 6 schizoaffective | 1977–1995 | 24 deaths, 21 deaths from natural causes, 2 suicides, 1 road accident, overall suicide rate was 1.3 per 1000 patient-years of observation |
Expected suicide rate per 1000 years in this population between 5.4 and 10.2 | ||||||
2001/Germany | Conell et al. | Affective disorders | Comparison of Li treatment with non-Li treatment period | 33 | Ø 10.3 years without lithium | Anti-suicidal effect did not differ after restarting the Li therapy |
Ø 19.8 years with lithium | ||||||
Response rate after discontinuation and restarting lower | ||||||
Retrospective cohort study | ||||||
2001/Germany (region Saxony) | Fülle et al. | All diagnoses | Analysis of suicide cases within all hospitals in Saxony compared to a matched control group | 800 patients including 400 suicides + 400 controls | Control group had 6 times more often lithium long-term treatment compared to suicide group | |
Lithium therapy was under-represented within the suicide group | ||||||
2001/Germany, Denmark, Austria | Ahrens et al. | Affective disorders | Li | 167 | Ø 6.7 years | Reduction of SA in excellent responders (3 SA), moderate responders (14 SA), and poor responders (21 SA) |
Retrospective cohort study | Subgroup of IGSLI cohort | 1120 treatment years | ||||
Comparison of response rates | ||||||
2001/USA | Coryell et al. | Affective disorders | Retrospective cohort study | 15 suicides compared to 15 non-suicides | 6 of suicides and 8 of controls were thought to take lithium, 9 of SA and 8 of controls were on lithium | |
41 suicide attempts compared to 41 non-SA | ||||||
2003/USA | Goodwin et al. | Bipolar disorder | Li | 20.623 | At least one prescription with Li or VLP or CBZ | Lithium-treated patients had 1.5- to 3-fold reduced suicide risk compared to VLP-treated patients |
VLP | Data from health insurances | |||||
CBZ | ||||||
Retrospective cohort study | ||||||
2005/Denmark | Kessing et al. | Li | 13.186 | Patients with at least one lithium prescription compared to patients who never received lithium | Lithium patients had 0.44-fold reduced suicide rate | |
Retrospective cohort study | Data from a national register | |||||
2005/Italy | Bocchetta et al. | Affective disorders | Li | 1394 patients | 5474 years of lithium treatment | Patients treated >5 years with lithium had reduced mortality rate (like general population) |
Retrospective cohort study | 18.154 patient years | |||||
2005/Switzerland | Angst et al. | Affective disorders | Li | 406 patients | 40 years | 45 suicides, Li-treated patients had lower suicide rate which did not differ from general population |
Prospective epidemiological study | ||||||
2006/Spain | Gonzales-Pinto et al. | Bipolar I | Li | 72 | 10 years | 5.2-fold reduced risk for suicidal behavior or SA in patients with stable and good Li response |
Prospective cohort study | ||||||
2008/USA | Collins and McFarland | Bipolar disorder | Li | 12662 patients | Overall 12 suicides, 81 SA, 2 suicides, 15 SA (Li) | |
VLP | 2558 (Li) | |||||
Other anti-convulsive medication | 2214 (VLP) | 2 suicides, 41 SA (VLP) | ||||
2002 (gabapentin) | 7 suicides, 19 SA (gabapentin) | |||||
242 (CBZ) | ||||||
4 SA (CBZ) | ||||||
Lowest (0.78) suicide rate per 1000 person years in Li-treated patients | ||||||
2009/Australia | Keks et al. | Bipolar disorder | Retrospective cohort study | 35 suicides | Only 4 of them were treated with lithium | |
2011/Germany | Neuner et al. | All diagnoses | Analysis of all suicides within psychiatric hospitals compared to a matched control group without suicides | 133 hospital suicides vs. 133 non-suicide patients | Affective disorders: 59 suicide patients (0 patients with lithium) 60 control patients (12 patients with lithium) |
Year | Author | Measurement | Number of samples | Results |
---|---|---|---|---|
2009/Japan | Ohgami et al. | Li level in drinking water | 18 municipalities | Standardized mortality ratio (SMR) negatively correlated with Li levels |
2011/UK | Kabacs et al. | Li level in drinking (tap) water | 47 samples from 47 subdivisions | No association between lithium levels in drinking (tap) water and mortality from suicide in the East of England |
2011/Austria | Kapusta et al. | Li level in drinking water | 6460 lithium measures of 99 Austrian districts | Suicide rate, SMR inversely associated with Li levels |
2013/Greece | Giotakus et al. | Li level in drinking water | 149 water samples from 34 prefectures | Tendency for lower suicide rates in the prefectures with high levels of lithium in drinking water |
2013/USA | Blüml et al. | Li level in public water | 3123 lithium water samples, 226 counties | Higher lithium levels in the public drinking water were associated with lower suicide rates |
2015/Italy | Vita et al. | Li level in drinking water | Review | Higher levels in drinking water may be associated with reduced risk of suicide in the general population |
Year/country | Author | Diagnoses | Medication | Number of patients | Duration | Results |
---|---|---|---|---|---|---|
1997/Italy, USA | Tondo et al. | Affective disorders | Comparison of Li treatment with non-Li treatment period | Over 17,000 | 8.6-fold higher risk for suicide and SA within the non-Li-treated group | |
After discontinuation of Li therapy, 7-fold increase of suicidal events | ||||||
Retrospective cohort study | ||||||
2005/UK | Cipriani et al. | Affective disorders | Li | 1389 Li-treated patients | Long-term treatment | Li reduced suicide risk and all-cause mortality by approximately 60 % |
Other compounds | ||||||
2069 with other-substances-treated patients | ||||||
2006/US | Baldessarini et al. | Affective disorders (mainly unipolar) | Li | 33,340 patients | 85.229 patient years | Lithium-treated patients had 5-fold reduced risk for suicides and SA compared to patients without lithium |
2007/USA, Italy | Guzzetta et al. | Unipolar depression | Li | 328 | 4.65 years with lithium | Overall risk for suicides and SA was 88.5 % lower with lithium |
6.27 years without lithium | ||||||
2013/UK | Cipriani et al. | Mood disorders | 15 comparisons | 6674 participants | 48 randomized controlled trials | Li more effective than placebo in reducing number of suicides and deaths from any causes |
Randomized controlled trials—not specifically focusing on suicidality or mortality
Studies focusing on suicidality and mortality: long-term follow-up of lithium-treated patients—epidemiological studies including record linkage
Studies on potential anti-suicidal effects of lithium as a trace element in drinking water
Reviews and meta-analyses
Discussion
Considerations for clinicians
Positive family history for bipolar disorder | |
Previous remission with lithium | |
Classical: euphoric manic episodes | |
Full remission between episodes | |
Good adherence |