Case Report: Bipolar 1 disorder

Review Suicide attempts in bipolar disorders:

comprehensive review of 101 reports Tondo L, Pompili M, Forte A, Baldessarini RJ. Suicide attempts in bipolar disorders: comprehensive review of 101 reports.

Objective: Assess reported risk of suicide attempts by patients with bipolar disorder (BD).

Method: Systematic searching yielded 101 reports from 22 countries (79 937 subjects). We analyzed for risk (%) and incidence rates (%/ year) of attempts, comparing sex and diagnostic types, including by meta-analysis.

Results: Attempt risk averaged 31.1% [CI: 27.9 –34.3] of subjects, or 4.24 [3.78 –4.70]%/year. In BD-I (43 studies) and BD-II subjects (30 studies), risks (29.9%, 31.4%) and incidence rates (4.01, 4.11%/year) were similar and not diﬀerent by meta-analysis. Among women vs.

men, risks (33.7% vs. 25.5%) and incidence (4.50 vs. 3.21%/year) were greater (also supported by meta-analysis: RR =1.35 [CI: 1.25 –1.45], P <0.0001). Neither measure was related to reporting year, % women/ study, or to onset or current age. Risks were greater with longer exposure, whereas incidence rates decreased with longer time at risk, possibly through ‘dilution’ by longer exposure.

Conclusion: This systematic update of international experience underscores high risks of suicide attempts among patients with BD (BD-I =BD-II; women >men). Future studies should routinely include exposure times and incidence rates by diagnostic type and sex for those who attempt suicide or not. L. Tondo 1,2,3 , M. Pompili 1,4 , A. Forte 1,4 , R. J. Baldessarini 1,2 1International Consortium for Bipolar & Psychotic Disorder Research, McLean Hospital, Belmont,2Department of Psychiatry, Harvard Medical School, Boston, MA, USA, 3Lucio Bini Mood Disorder Centers, Cagliari and Rome, 4NESMOS, Sant ’Andrea Medical Center, La Sapienza University of Rome, Rome, Italy Key words: attempt; bipolar disorder; onset age; rates; suicide Dr. Leonardo Tondo, McLean Hospital-Harvard Medical School, MRC3, 115 Mill Street, Belmont, MA 02478,USA. E-mail: [email protected] Accepted for publication September 29, 2015 Summations Suicide was attempted at least once by 31.1% of bipolar disorder subjects, or 4.24% of persons/year.

Suicide attempters were 35% more frequent among women than men.

There was no diﬀerence in suicide attempter risk between types I and II bipolar disorder. Considerations Higher frequency of suicide attempts in women and no diﬀerence in bipolar diagnostic subtypes con- ﬁrm previous ﬁndings.

Small sample size and brief exposure times can exaggerate estimates of suicidal risk.

We recommend reporting suicidal risk with correction for exposure time (incidence rate), as %/year. Introduction Suicidal behaviour is a major, international public health challenge (1 –4). Mood disorders are associ- ated with especially high standardized mortality ratios (SMRs) for suicide compared with the gen- eral population (3 –5). Bipolar disorder (BD) accounts for the highest SMR for suicide, followed by major depression severe enough to lead to psy- chiatric hospitalization, with lower risks among outpatient with less severe depressive disorders or patients with most other psychiatric diagnoses (6 – 10). Risk factors for suicide and attempts have been considered in recent reviews and comprehen- sive treatises on BD or suicide (2, 11 –13). In BD, suicide is particularly associated with recurrent 174 Acta Psychiatr Scand 2016: 133: 174–186 ©2015 John Wiley & Sons A/S. Published by John Wiley & Sons Ltd All rights reserved DOI: 10.1111/acps.12517 ACTA PSYCHIATRICA SCANDINAVICA and acute depressive phases and even more likely during or even following mixed, agitated-dysphoric states (6, 7, 14). However, the strongest identiﬁed risk factor for suicide generally is a previous sui- cide attempt (15, 16). Notably, self-harm has been associated with increased all-cause mortality (17).

The quantitative relationship of suicide attempts to suicides among psychiatric patients diﬀers greatly from the general population (18–20).

Reported rates of attempts in the general popula- tion have averaged 0.2–0.6% per year, with an average international suicide rate of approximately 0.01%/year, for a ratio of attempts/suicides (A/S) of approximately 40 (0.4/0.01) (18, 19). The A/S ratio is a suggestedindex of lethalityof suicidal acts (10, 19). In contrast to the high ratio of attempts to suicides in the general population, this ratio was much lower in both type I (with mania) and II (with hypomania) patients with BD, averag- ing less than 10 (7) to indicate much greater lethal- ity of attempts among patients with BD than in the general population (10, 19). In the general popula- tion, risk of death by suicide is particularly high among young, relatively impulsive-aggressive men and older, unmarried or socially isolated men, whereas women are more than twice as likely as men to attempt suicide (2, 3, 18, 21, 22). These relationships for women and men diagnosed with BD are less certain, particularly owing to a lack of secure information about rates of suicide attempts in samples of patients with BD distinguished by sex and corrected for exposure times at risk (8, 9).

Despite a growing range of clinically eﬀective treatments available for mood disorders, suicide risk in such patients, and especially those with BD, remains high, probably reﬂecting the great diﬃ- culty of eﬀectively treating depressive and mixed states in BD, and general diﬃculties of maintaining consistent adherence to any long-term treatment (8, 14, 23, 24). Such enduring suicidal risk proba- bly reﬂects the high long-term proportion of weeks ill among clinically treated patients with BD. This proportion has averaged 40–50%, even from ill- ness onset, and fully three-quarters of this morbid- ity were depressive or dysphoric (25–29). Surely, such unresolved depressive morbidity contributes importantly to risk of suicide and other adverse or even fatal outcomes (30, 31).

Reported proportions of women and men diag- nosed with type I or II BD who make suicide attempts vary considerably. Novick et al. (8) reviewed risks of suicide attempts in BD-I and BD- II patients from 24 randomized, controlled trials involving 4899 subjects, with meta-analyses based on 15 of the studies, and found similar average risks, of 36.3% in BD-I and 32.4% in BD-IIdisorder subjects (OR=1.21, 95% CI: 0.98–1.48, P=0.07). More recently, Schaﬀer et al. (9) in an international task force reviewed demographic and clinical correlates of attempted and completed sui- cide in patients with BD, based on 24 reports involving over 44 000 BD subjects. Notable risk factors associated with suicide attempts in patients with BD included: female sex, younger age at onset, depressive ﬁrst-lifetime episode as well as depressive current or most recent episodes, co- occurring anxiety or substance-use disorders, bor- derline personality characteristics, and suicide in a ﬁrst-degree relative. In 14 meta-analyzed studies (9), they found suicide attempts to be more fre- quent in BD-I vs. BD-II patients in two, more in BD-II in two others, not diﬀerent in the remaining 10, and indistinguishable by diagnostic type overall (OR=1.07, CI: 0.79–1.45,P=0.68). They also found a highly signiﬁcantly higher risk of suicide attempts among women (OR=1.54, CI: 1.44– 1.66,P<0.0001).

Despite these advances, our impression was that there is additional information concerning risks of suicide attempts, particularly to quantitate possible diﬀerences between diagnostic subtypes and the sexes, as well as a need to consider eﬀects of time at risk. Accordingly, we undertook the present, systematic, and quantitative review of available research ﬁndings. Aims of the study With a systematic review of reported risks (%) of suicidal acts in BD, we aimed primarily to establish overall risks, and by diagnostic type (I vs. II), in women vs. men, and for eﬀects of exposure time, as secondary aims. We also hypothesized that longer exposure would increase observed preva- lence (% of persons aﬀected) but tend to reduce the incidence rate (% aﬀected per year, or of per- sons attempting suicide per person-year).

Material and methods Eligibility criteria and information sources We used computerized literature searching to iden- tify reports appearing between 1976 and March 2015, containing data on suicide attempts in type I or II patients with BD, as well as information on times of observation or on risk exposure, or esti- mates of onset age and age at intake suitable to assess average years at risk. We searched the MEDLINE/PubMed , PsychINFO , and EMB ASE research literature databases with the fol- lowing terms: ‘bipolar disorder’ AND ‘suicide 175 Suicide attempts in bipolar disorders attempt’. In addition, we hand-searched the biblio- graphy of screened reports and of reviews for addi- tional citations. Exclusion criteria were as follows:

failure to report on proportion of persons with sui- cide attempts, diagnosis other than BD or without separation from major depression or other diag- nostic groups, reports only on completed suicides, or on suicide attempts but not on persons who attempted, sampling only from pediatric subjects, repeated reports from the same study, case reports, and not reporting in English. In addition, we excluded studies reporting on samples selectively receiving long-term treatment with lithium as it is associated with lower suicidal risk (10), but included samples with phases of treatment not involving lithium or with unstated treatments.

Moreover, studies were excluded if they reported only on suicidal ideation, apparently non-suicidal self-injury, or if suicidal status was rated as a con- tinuous measure based on estimated severity.

Search strategy Searching initially yielded 293 records, abstracts of which were screened (by LT and MP), leading to exclusion of 170 publications not relevant to this study, reviews, or duplicate reports. The remaining 123 abstracts were reviewed independently by all investigators; disagreements were resolved by con- sensus, and 22 records were excluded for not meet- ing study criteria. This process yielded 101 full texts to be assessed in detail for eligibility, all of which were included in the reported analyses (7, 15, 32–130).

Data extraction and analysis All four authors extracted data regarding the fol- lowing: (i) proportions of persons with suicide attempts (prevalence or risk) during reported expo- sures times, distinguishing by sex and diagnostic type (BD-I, BD-II) when possible; (ii) average exposure time either reported explicitly as time under observation (27 studies) or estimated as half of the diﬀerence between reported ages at intake and at onset (74 studies). Any disagreements were resolved by consensus.

Risk of persons with at least one suicide attempt was calculated as a proportion (%) of the total sample considered. Data are presented as means standard deviations (SD) or with 95% conﬁdence intervals (CI). Incidence rates (%/year) were the ratios of the proportion of persons with suicide attempts to estimated years at risk, to adjust for varying risk exposure times. Both risk estimates were further analyzed by regressing themvs. available individual factors of interest: publica- tion year, sample size, ages at onset and intake, and average exposure times, to provide linear (r) or nonparametric (r s) correlations. We also ana- lyzed risks and rates by multivariable linear regres- sion modeling and used random eﬀects meta- analytic methods to compare risks and rates by sex and by diagnosis, using the stated potential covari- ates in both procedures. Analyses were based on STATVIEW .5 (SAS Institute, Cary, NS, USA) spreadsheets and STATA .12.1 (StataCorp, College Station, TX, USA) commercial statistical software. Results Overall findings The 101 included studies (7, 15, 32–130; Table 1) reported between 1976 and 2015 provided infor- mation on persons attempting suicide among 79 937 BD subjects from 22 countries: USA (n=37), Italy (n=15), France (n=8), Brazil and Spain (n=6), Canada (n=4), Japan and Sweden (n=3), two each from Czech Republic, Finland, New Zealand, South Korea, and Switzerland, and one each from Australia, China, Ethiopia, Hun- gary, Iran, the Netherlands, Norway, Taiwan, and Turkey. In studies providing the required data, 62.9% of subjects were women (25 studies), and 81.9% were diagnosed with BD-I (BD-I, 43; BD- II, 30 studies). Onset age averaged 24.2 [CI: 23.4– 25.0] years and age at study intake averaged 41.6 [40.6–42.7] years. Time at risk averaged 8.00 [7.38– 8.62] years.

A total of 24 860 of the 79 937 subjects were reported to have had at least one suicide attempt, providing an overall risk (prevalence) of 31.1% [27.9–34.3] (Table 1). The annualized incidence rate (percentage of persons with at least one sui- cide attempt per year at risk) averaged 4.24% [3.78–4.70]%/year (Table 1). These risks and rates were highly correlated, as expected (r s=0.639, P<0.0001), but neither was related to the year of reporting (bothr≤0.10;P≥0.15). There was no signiﬁcant diﬀerence in mean annualized rates in 80 studies yielding estimates of exposure as half of years ill (4.17 [CI: 3.70–4.64]%/year) and 21 reporting actual time of observation (4.50 [3.05– 5.95]%/year) supporting the validity of the method used in calculating exposure time. Comparisons of risks by sex and by diagnostic type Comparisons of suicidal risks in women vs. men were based on 25 reports that included such data for both sexes (Table 2). Based on 48 245 subjects, 176 Tondo et al. Table 1. Summary of 101 international reports of suicide attempts in bipolar disorder patients Study CountryYears Attempts (n) Subjects (N)Attempt risk and incidence rate Onset Age Exposure % %/year Dunner et al. (32) USA 34.0 46.0 6.00 20 45 44.4 7.41 Hanus & Zapletalek (33) Czech Rep. 24.2 41.6 5.00 29 99 29.3 5.86 Coryell et al. (34) USA 24.9 37.0 6.05 14 69 20.3 3.35 Roy-Byrne et al. (35) USA 28.0 39.8 5.90 39 67 58.2 9.87 Wehr et al. (36) USA 30.0 40.0 5.00 32 70 45.7 9.14 Rihmer et al. (37) Hungary 24.2 41.6 7.20 25 36 69.4 9.65 Wu and Dunner (38) USA 22.0 42.0 10.0 72 220 32.7 3.27 Nordstr€ om et al. (15) Sweden 27.0 49.8 7.00 8 55 14.6 2.08 Chen & Dilsaver (39) USA 20.7 46.5 12.9 49 168 29.2 2.26 Feinman & Dunner (40) USA 22.1 41.6 9.72 72 188 38.3 3.94 Bocchetta et al. (41) Italy 26.4 38.1 5.85 100 1116 8.96 1.53 Perugi et al. (42) Italy 26.3 35.0 4.35 69 261 26.4 6.08 Vieta et al. (43) Spain 26.6 37.7 5.55 18 60 30.0 5.41 Goldberg et al. (44) USA 26.7 40.0 6.65 82 184 44.6 6.70 Cassano et al. (45) Italy 31.0 45.7 7.35 26 129 20.2 2.74 Tondo et al. (46) Italy 29.4 40.8 8.28 46 310 14.8 1.79 Tsai et al. (47) Taiwan 22.7 44.0 9.50 53 101 52.5 5.52 Perugi et al. (48) Italy 26.8 37.8 5.50 83 320 25.9 4.72 Henry et al. (49) France 24.2 42.7 2.67 4 44 9.09 3.40 Suppes et al. (50) USA 22.9 43.9 10.5 75 261 28.7 2.74 Vieta et al. (51) Spain 25.5 40.6 7.55 35 129 27.1 3.59 Dittmann et al. (52) USA 24.4 42.1 8.83 53 141 37.6 4.27 Rucci et al. (53) USA 22.0 35.1 6.55 50 175 28.6 4.36 Serretti et al. (54) Italy 32.9 48.1 7.60 177 625 28.3 3.73 Ten Have et al. (55) the Netherlands 26.2 35.1 4.45 22 136 16.2 3.64 Bowden et al. (56) USA 23.4 40.7 8.65 102 347 29.4 3.40 Calabrese et al. (57) USA 22.7 42.2 9.75 353 958 36.8 3.78 Cate-Carter et al. (58) Canada 20.3 35.4 7.55 80 320 25.0 3.31 Goodwin et al. (59) USA 24.2 41.6 1.00 92 2413 3.81 3.81 Henry et al. (60) France 27.0 53.3 13.2 127 318 39.9 3.04 Judd & Akiskal (61) USA 26.0 42.6 8.30 58 233 24.9 3.00 Leverich et al. (62) USA 19.7 41.4 10.5 125 397 31.5 3.00 MacKinnon et al. (63) USA 20.0 40.8 10.4 202 603 33.5 3.22 Sakamoto & Fukunaga (64) Japan 31.1 39.5 4.22 16 80 20.0 4.73 Yerevanian et al. (65) USA 24.2 41.6 11.5 13 140 9.29 0.81 Dalton et al. (66) Canada 20.1 35.4 7.63 86 336 25.6 3.35 Engstrom et al. (67) Sweden 28.0 54.0 13.0 23 100 23.0 1.77 Ernst & Golberg (68) USA 21.7 40.8 9.55 17 40 42.5 4.45 Fagiolini et al. (69) USA 22.2 35.2 6.00 50 175 28.6 4.76 Joyce et al. (70) New Zealand 17.3 30.7 6.70 9 19 47.4 7.07 Perlis et al. (71) USA 17.4 40.6 11.6 352 983 35.8 3.09 Raja & Azzoni (72) Italy 24.2 47.7 5.80 31 80 38.8 6.68 Ran et al. (73) China 29.0 37.0 4.00 8 47 17.0 4.26 Shi et al. (74) USA 24.2 44.1 1.00 25 5379 0.46 0.46 Slama et al. (75) France 24.2 44.2 3.50 109 307 35.5 10.1 Vieta et al. (76) Spain 24.2 40.2 8.05 13 107 12.2 1.51 Angst et al. (77) Switzerland 24.2 73.2 11.0 79 220 35.9 3.26 Hajek et al. (78) Czech Rep. 24.2 44.0 9.99 64 242 26.4 2.65 Kawa et al. (79) New Zealand 19.0 43.5 12.2 95 211 45.0 3.68 Kupka et al. (80) USA 20.4 42.5 11.1 55 539 10.2 0.92 Moreno & Andrade (81) Brazil 24.2 38.8 7.39 5 21 23.8 3.22 Valenc a et al. (82) Brazil 15.2 42.6 13.7 9 41 22.0 1.60 Weiss et al. (83) USA 17.5 40.6 11.6 315 1000 31.5 2.73 Galfalvy et al. (84) USA 22.6 42.5 9.95 12 64 18.8 1.88 Gonzalez-Pinto et al. (85) Spain 28.0 51.8 11.9 19 72 26.4 2.22 Grunebaum et al. (86) USA 24.2 40.7 8.25 64 96 66.7 8.08 Valtonen et al. (87) Finland 24.2 37.9 1.50 29 160 18.1 12.1 Bader et al. (88) USA 18.0 43.0 12.5 53 175 30.3 2.42 Gazalle et al. (89) Brazil 24.8 42.5 8.82 82 169 48.5 5.50 Lopez et al. (90) USA 19.0 41.7 11.4 470 1255 37.4 3.30 Simon et al. (91) USA 24.2 38.4 2.70 1719 32 360 5.31 1.97 Simon et al. (92) USA 17.3 44.2 13.4 46 120 38.3 2.85 Tondo et al. (7) Italy 29.7 37.9 4.10 66 517 12.8 3.11 177 Suicide attempts in bipolar disorders the risk of attempts in women averaged 33.7% [CI:

27.3–40.2], and in men, 25.5% [19.6–31.3] (a 32.2% diﬀerence; pairedt-test=11.4,P<0.0001).

The respective annualized rates were 4.50%/year [3.48–5.52] in women and 3.21%/year [2.68–3.74] in men (a 1.40-fold diﬀerence; pairedt-test=3.44, P=0.002).

Comparisons of suicidal risks in persons with BD-I vs. BD-II were based on 22 388 subjects from 43 and 30 reports respectively (Table 2).

There was no signiﬁcant diﬀerence in risk of sui- cide attempts between these diagnostic groups:

29.9% [CI: 25.9–34.0] in BD-I subjects and 31.4% [24.5–37.4] in BD-II—5.02% higher in BD-II (t=0.16,P=0.87). Incidence rates also were not signiﬁcantly diﬀerent: 4.01%/year [3.48–4.54] in BD-I and 4.11%/year [3.23–4.99] in BD-II— 2.49% higher among BD-II subjects (t=0.35, P=0.73). There were insuﬃcient reports (n=3–5)to support credible analyses of risks or rates by sex among BD-I vs. BD-II diagnostic groups. Effects of other factors Regression of risk of attempting suicide vs. age at onset indicated a modestly higher risk (% of sub- jects) with younger onset age (r= 0.22, P=0.02), but the incidence of suicide attempters (%/year) was not signiﬁcantly related to onset age (r=0.10,P=0.32). In addition, risk (%) was not signiﬁcantly associated with age at intake (r=0.03,P=0.78), whereas higher incidence (%/ year) was signiﬁcantly associated with younger current age (r= 0.30,P=0.002). However, these preliminary relationships of onset age or current age with risk or rate of suicide attempts were not signiﬁcant following adjustment for exposure time, based on multivariable linear regression modeling. Table 1. (Continued) Study CountryYears Attempts (n) Subjects (N)Attempt risk and incidence rate Onset Age Exposure % %/year Azorin et al. (93) France 27.3 45.5 9.10 366 958 38.2 4.20 Carballo et al. (94) USA 18.1 35.1 8.50 116 148 78.4 9.22 Cruz et al. (95) Spain 30.0 44.7 7.35 211 3089 6.83 0.93 Gao et al. (96) USA 24.2 35.6 5.80 109 245 44.5 7.67 Nakagawa et al. (97) Japan 22.4 38.5 8.03 68 116 58.6 7.30 Pompili et al. (98) Italy 24.2 46.5 11.2 28 61 45.9 4.08 Raja et al. (99) Italy 24.2 43.5 3.00 85 738 11.5 3.84 Sajatovic et al. (100) USA 21.8 45.8 12.0 177 277 63.9 5.32 Sublette et al. (101) USA 18.8 37.4 9.30 105 138 76.1 8.18 Azorin et al. (102) France 28.0 43.2 7.60 382 1090 35.0 4.61 Malloy-Diniz et al. (103) Brazil 19.2 40.9 10.8 18 39 46.2 4.25 Raja et al. (104) Italy 24.2 43.5 9.70 85 823 10.3 1.06 S anchez-Gistau et al. (105) Spain 26.3 41.5 7.60 112 390 28.7 3.78 Monteleone et al. (106) Italy 27.8 48.0 10.1 17 194 8.76 0.87 Ryu et al. (107) So. Korea 28.5 35.7 3.60 76 579 13.1 3.65 Bellivier et al. (108) France 28.7 44.7 8.00 663 2219 29.9 3.73 Dennehy et al. (109) USA 16.0 37.0 1.31 174 4352 4.00 3.05 Mitchell et al. (110) Australia 20.0 48.0 14.0 80 246 32.5 2.32 Perroud et al. (111) Switzerland 24.2 41.1 8.50 67 138 48.6 5.71 Bega et al. (112) Canada 24.3 37.1 6.40 342 1429 23.9 3.74 Finseth et al. (113) Norway 19.7 42.3 11.3 93 206 45.2 4.00 Parmentier et al. (114) France 25.4 40.2 7.40 280 652 42.9 5.80 Antypa et al. (115) Italy 16.1 40.5 12.2 1023 2884 35.5 2.91 Bani-Fatemi et al. (116) Canada 21.8 35.6 6.90 74 341 21.7 3.15 Shabani et al. (117) Iran 25.9 35.2 4.65 33 100 33.0 7.10 Benedetti et al. (118) Italy 27.5 45.8 9.15 32 136 23.5 2.57 Holma et al. (119) Finland 28.1 39.2 5.55 82 147 55.8 10.0 Kurumaji et al. (120) Japan 34.2 46.0 5.90 48 128 37.5 6.36 Nery et al. (121) Brazil 25.9 40.9 7.50 147 374 39.3 5.24 Sandberg et al. (122) Sweden 24.2 41.6 1.00 8 98 8.16 8.16 Shibre et al. (123) Ethiopia 24.2 41.6 10.0 63 263 23.9 2.40 Zimmerman et al. (124) USA 24.2 38.6 7.25 96 263 36.5 5.03 Da Grac a et al. (125) Brazil 24.2 41.6 14.0 37 66 56.1 4.00 G€ ucl€ u et al. (126) Turkey 24.2 35.3 10.3 34 96 35.4 3.44 Kim et al. (127) So. Korea 24.2 34.7 11.3 121 512 26.3 2.09 Oli e et al. (128) France 22.0 41.8 9.90 129 343 37.6 3.80 Tundo et al. (129) Italy 32.5 45.3 6.40 20 101 19.8 3.09 Zeng et al. (130) USA 24.2 36.7 6.30 11 54 20.4 3.23 Mean or total [95% CI] 101 countries 24.2 [23.4–25.0] 41.6 [40.5–42.7] 8.00 [7.38–8.62] 12 032 79 937 31.1 [27.9–34.3] 4.24 [3.78–4.70] 178 Tondo et al. Time at risk (either reported or averaged between ages at onset and at intake) was strongly associated with greater risk (r=+0.35, P=0.0003) but lower incidence rate (r= 0.38, P<0.0001; Fig. 1). Signiﬁcant associations also were found between smaller sample size and larger apparent prevalence (r= 0.24,P=0.02) and incidence rate (r= 0.29,P=0.003). However, these signiﬁcant correlations disappeared when four unusually large studies (74, 91, 95, 109) with more than 3000 subjects each were excluded in a sensitivity re-analysis. Multivariable modeling Multivariable regression modeling used overall risks and rates as the dependent variables. It found an independent and signiﬁcant association only with exposure time, which was positive for risk [slope coeﬃcient (b)=+2.17,P<0.0001] and neg- ative for rate (b= 0.024,P=0.006), similar tothe results of simpler regression models. That is, risk or prevalence of suicide attempt increased with exposure, as expected, whereas the annualized inci- dence ratedecreased. Other variables (publication year, ages at onset and at intake, sample size, and per cent of women and of BD-I subjects) were not associated with the outcome in multivariable mod- eling. Meta-analysis We used random eﬀects meta-analysis to compare risk for attempted suicide in men vs. women and in BD-I vs. BD-II disorder subjects. Based on 25 studies providing data for paired comparison of women vs. men, the pooled risk ratio (RR) was (1.35 [CI: 1.25–1.45],z=8.14,P<0.0001), indi- cating a highly signiﬁcant, 35% greater risk among women with either type of BD (Fig. 2). Similar to the lack of diﬀerences in mean risks or rates, the meta-analytic comparison of BD-I and BD-II sub- jects in 27 reports, resulted in a non-signiﬁcant pooled risk ratio (RR=1.03 [0.88–1.20],z=0.37, P=0.71; Fig. 3). Meta-regression modeling indi- cated that none of the covariates available for test- ing (study size, publication year, onset and current age, or years at risk, and either sex or diagnosis) was associated with sex or diagnosis. Applying funnel plot and Eggers’ tests there was no evidence of signiﬁcant reporting bias, such as by selective reporting of ﬁndings with relatively high suicide risks.

Lethality of suicide attempts We also considered the ratio of attempts to sui- cides (A/S) among patients with BD as a proposed index of lethality (16). The present ﬁndings indi- cate an average incidence of suicide attempters among patients with BD of 4.24%/year (Table 1). Table 2. Summary of suicide attempt data in bipolar disorder patients Groups Reports (n) Subjects (n)Attempters Risk/100 persons)Rate/100 person-years) All subjects 101 79 937 31.1 [27.9–34.3] 4.24 [3.78–4.70] Sex Women 25 30 366 33.7 [27.3–40.2]* 4.50 [3.48–5.52]† Men 25 17 879 25.5 [19.6–31.3]* 3.21 [2.68–3.74]† Diagnosis Bipolar I 43 18 565 29.9 [25.9–34.0]‡4.01 [3.48–4.54]§ Bipolar II 30 3738 31.4 [24.5–37.4]‡4.11 [3.23–4.99]§ Computations are based on averages of risks and exposure times in individual studies.

Based on pairedt-tests:

*All subjects, women vs. men (%):t=11.4,P<0.0001.

†All subjects, women vs. men (%/year):t=3.44,P=0.002.

‡bipolar I vs. II (%):t=0.16,P=0.87.

§bipolar I vs. II (%/year):t=0.35,P=0.73. 0 10 20 30 40 50 60 70 80 Attempt Risk (% of subjects) 02468101214 Exposure (years at risk) 0 1 2 3 4 5 6 7 8 910 11 12 Attempt Rate (%/year) 02468101214 r= +0.353, P= 0.0003r= P < 0.0001 (a) (b) Fig. 1.(a) Risk (prevalence; % of subjects) of suicide attempt among bipolar disorder patients vs. years-at- risk. (b) Rate (%/year) of suicide attempts vs. years-at-risk. Note that both functions are highly signiﬁcant but opposite in direction (a)r=+0.362, P=0.0003; (b)r= 0.383, P<0.0001). 179 Suicide attempts in bipolar disorders The incidence rate of suicides in patients with BD is reported to average 0.36%/year: 0.44 in men, 0.28 in women (9, 131). Therefore, the extrapo- lated A/S ratio in patients with BD is (4.24/0.36), or 11.8 overall, and is lower in men (3.21/ 0.44=7.30) than in women (4.50/0.28=16.1), indicating 2.21-fold greater apparent lethality of attempts among men. These ratios are several times lower than has been observed in the general population, indicating much greater overall lethal- ity of attempts in patients with BD (10, 18).

Discussion The mean risk or prevalence of attempting suicide in this systematic review involving nearly 80 000 BD subjects in 101 studies from 22 countries (1976–2015) was 31.1% [27.9–34.3%], with very similar risks in types I and II BD (Table 2; Fig. 3).

As the risk was strongly inﬂuenced by exposure time, we also standardized risks to incidence rates per year. The adjusted incidence indicated a suicide attempt in about four persons diagnosed with BDper 100 cases in each of the average of 8.00 years of exposure represented in the available data (4.24 [3.78–4.70]%/year). The years at risk were from an average onset age of 24 years, to age of observa- tion averaging 42 years. For the incidence rate of suicidal persons per year, there was again no diﬀer- ence between BD-I and BD-II subjects, but it was signiﬁcantly higher among women than men (Table 2)—35% higher by meta-analysis (Fig. 2).

That ﬁnding and the lack of diﬀerence in risk between types I and II BD conﬁrm those of previ- ous reports (8, 9, 16). The risk and incidence of sui- cide attempts reported here are representative of recent international experience that for the most part involved risks in the presence of clinical treat- ment (other than speciﬁc use of lithium).

Risk of suicide attempts also was weakly associ- ated with younger onset age. It is not clear that this association indicates that earlier onset may be associated with more severe, or suicide prone, ill- ness in BD, as has been suggested previously (13, 132, 133), or may merely reﬂect a higher probabil- ity of suicidal acts with more years of illness, as, as Female Male 9/22 11/71 33/100 30/84 19/74 7/55 26/198 15/112 54/207 32/119 136/374 83/274 19/60 4/40 57/121 38/90 19/32 18/32 20/91 7/85 62/122 20/47 1269/21200 450/11160 31/71 15/49 45/317 21/152 249/628 133/462 69/213 43/177 46/317 30/262 436/1226 206/932 118/2477 56/1875 56/113 37/93 176/378104/274 675/1674 348/1210 21/93 11/43 39/63 11/23 88/195 41/148 Pooled 3772/30366 1771/17769 –2–1012345 Relative risk (RR [95%CI]) Pooled Olie et al. 2015 [128] DaGraca et al. 2015 [125] Benedetti et al. 2014 [118]Antypa et al. 2013 [115] Parmentier et al. 2012 [114]Finseth et al. 2012 [43] Dennehy et al. 2011 [109]Bellivier et al. 2011 [108]Ryu et al. 2010 [107] Sanchez-Gistau et al. 2009 [105]Azorin et al. 2009 [93]Tondo et al. 2007 [7] Simon N et al. 2007 [92] Simon G et al. 2007 [91]Gazalle et al. 2007 [89] Valtonen et al. 2006 [87]Galfalvy et al. 2006 [84]Kawa et al. 2005 [79] Engstrom et al. 2004 [67]Leverich et al. 2003 [62]Dalton et al. 2003 [66]Tondo et al. 1999 [46] Cassano et al. 1999 [45] Goldberg et al. 1998 [44]Dunner et al. 1976 [32] more in women null Pooled RR = 1.35 [1.25–1.45]P<0.0001 Fig. 2.Forest plot of random-eﬀects meta-analytic ﬁndings from 25 trials comparing risk of suicide attempts in bipolar disorder by sex.

On the right are proportions of subjects making a suicide attempt/all subjects, among women vs. men. The pooled risk ratio (RR) is highly signiﬁcant (RR = 1.35 [95% CI: 1.25–1.45]; z = 8.14,P< 0.0001) and (RR = 1.03 [95% CI: 0.88–1.20]; z = 0.37,P= 0.71).

180 Tondo et al. expected, exposure time was strongly associated with a higher proportion of subjects attempting suicide (Fig. 1a). Of note, in addition to low rates of reporting on suicidal risks by both sex and diag- nostic type (<5% of studies reviewed), we found only limited information concerning many demo- graphic and clinical characteristics of subjects which would be of interest for relationships to sui- cidal behaviour, as noted by other reviewers (13, 16).

Risks and incidence rates of suicide attempts were not associated with the year of publication nor was there evidence of bias toward reporting higher risks of attempts. However, sample size was signiﬁcantly, negatively, associated with suicide attempts to yield larger risks with smaller samples.

Exclusion of lithium-treated populations was justi- ﬁed by the possible role of this agent to lower the rate of suicide attempts (10, 19). In a previous study, we found a rate of suicidal acts (includingsuicides) in lithium-treated patients of 1.08%/year, or only one-fourth that observed in the present ﬁndings, whereas the corresponding rate among patients with BD not treated with lithium was 3.63%, or very similar to that reported here (19).

It is also interesting to consider the ratio of attempts to suicides (A/S) among patients with BD as a proposed index of lethality (10, 19). We esti- mated an A/S ratio in patients with BD of 11.8 overall, 7.30 in men, and 16.1 in women, compared to an estimated A/S ratio of at least 100 for women, and 21 for men in the general population (10, 15, 132). The relatively low A/S ratio associ- ated with BD appears to indicate that the intent and the lethality of means associated with suicide attempts among patients with BD averages several times greater than in the general population, par- ticularly among women with BD vs. the general population. That is, the A/S ratio, compared with the general population, is approximately 6.2 times BD-I BD-II 11/29 9/16 7/29 7/40 12/38 6/22 9/35 17/94 34/353 7/25 40/107 1336 123/489 54/136 61/229 19/91 98/305 20/68 12/40 4/40 19/75 4/25 52/160 27/60 3/14 2/7 20/37 17/27 11/81 16/95 24/67 21/72 439/1158 31/97 28/75 12/30 45/326 21/191 21/40 7/21 75/96 30/42 72/234 41/156 128/2797 35/1166 270/935 72/494 60/140 33/66 204 86/308 62/184 51/125 Totals 1771/8277 662/3550 –1 0 1 2 3 4 5 Pooled Olie et al. 2014 [128]Kim et al. 2015 [127] Finseth et al. 2012 [113]Bega et al. 2012 [112] Dennehy et al 2011 [109] Sanchez-Gistau et al. 2009 [105]Sublette et al. 2008 [101]Pompili et al. 2008 [98]Tondo et al. 2007 [7] Simon N et al. 2007 [92]Lopez et al. 2007 [90] Bader et al. 2007 [88] Valtonen et al. 2006 [87]Galfalvy et al. 2006 [84] Moreno & Andrade 2005 [81]Angst et al. 2005 [77] Engstrom et al. 2004 [67] Sakamoto & Fukunaga 2003 [64]Leverich et al. 2003 [62]Dalton et al. 2003 [66] Serretti et al 2002 [54] Dittmann et al 2002 [52]Tondo et al. 1999 [46] Cassano et al 1999 [45]Vieta et al 1997 [43] Coryell et al 1987 [34] Dunner et al 1976 [32] Relative risk (RR [95%CI]) Pooled RR = 1.03 [0.88 –1.20] P =0.71 more in BD-I null Fig. 3.Forest plot of random-eﬀects meta-analytic ﬁndings from 27 trials comparing risk of suicide attempts in types I vs. II bipolar disorder. Data on the right are numbers of subjects making a suicide attempt/all subjects, for BD-I vs. BD-II subjects. The pooled risk ratio (RR) does not diﬀer from the null value of 1.00 (RR=1.03 [95% CI: 0.88–1.20;z=0.37,P=0.71). 181 Suicide attempts in bipolar disorders (100/16.1) lower in women with BD, compared with 2.9-fold in men with BD (21/7.30) (10, 15, 134). It may also be signiﬁcant that the A/S ratio has been found to be about twice greater (lesser lethality) among patients with BD when treated with lithium than without it (10, 19). These consid- erations indicate that the A/S ratio is lower (greater lethality) among men than women diag- nosed with BD. However, our eﬀorts to compare women and men in the two diagnostic subtypes, speciﬁcally, were severely constrained by the pau- city of studies of sex diﬀerences in types I and II BD (Table 2).

In general, the incidence rate measure of risk adjusted for exposure time (%/year) is a more suit- able representation of the risk of suicide attempts as the proportion of patients identiﬁed as aﬀected (%) was strongly increased by longer exposure time (Fig. 1a). Notably, however, the incidence rate (%/year) of attempts stronglydecreasedwith longer observation times, suggesting that studies involving short periods at risk may well overesti- mate attempt rates. We further propose that the seemingly paradoxical decline of attemptratewith longer exposure time probably may represent ‘dilu- tion by time’. Such an eﬀect would be expected in a non-chronic or episodic disorder like BD and has been reported previously as regards rates of epi- sodes of illness per year (135–137). In addition, we found that large samples lowered observed risks of suicide attempts in BD subjects, perhaps because infrequent events may be overlooked more often.

When such studies were omitted, the relationship of lower observed risk of suicide attempts was no longer signiﬁcant.

Limitations A notable limitation of this study concerns the apparently incidental reporting of suicide attempts in many of the reports analyzed which may lead to inaccurate identiﬁcation of suicidal acts. In addi- tion, observed suicidal incidence rates may be overestimated in small samples and with short periods of risk exposure. An additional limitation is the infrequent reporting separate rates for important subgroups, including by sex and diag- nostic type. Moreover, the pooled results are likely to be aﬀected by diﬀerences in clinical history and treatments, which, in addition to exposure times, are not included in most of the available reports.

Final considerations This study provides the largest, systematic sum- mary of reported risks of suicide attempts in sam-ples of subjects diagnosed with a BD, including comparisons of types I and II BD and of women vs. men. The ﬁndings underscore the importance of assessing risk for suicide attempts with consider- ation of exposure time, given strong, inverse, dependence of risk (% of persons with attempts) with exposure time at risk. We recommend for future reports that average exposure times or years of illness be reported routinely, with risks adjusted for exposure time and that sex and diagnostic types be provided for those who attempted suicide and not only for the overall sample considered. The ﬁndings support the impression that suicidal beha- viour is a particularly compelling public health and clinical problem in BD and that the risks are strik- ingly similar in types I and II BD. Acknowledgements Supported in part by an award from the Aretæus Foundation of Rome and the Centro Bini Private Donors Research Fund (to LT), and by the Bruce J Anderson Foundation and the McLean Private Donors Research Fund (to RJB).

Declaration of interest No author or any immediate family member has ﬁnancial rela- tionships with commercial or industrial entities that might rep- resent potential conﬂict of interest in this work.

References 1.GoldsmithSK,PellmarTC,KleinmanAM,BunneyWEJ.

Reducing suicide. Washington, DC: Institute of Medicine of the US National Academies of Science, 2002.

2.HawtonK, vanHeeringenK. Suicide. Lancet 2009;373:1372–1381.

3.SimonRI,HalesRE. Textbook of suicide assessment and management, 2nd edn. Washington, DC: American Psy- chiatric Press; 2012.

4. World Health Organization. Preventing suicide: a global imperative. Geneva World Heal Organ Published Online First, 2014, http://www.who.int/mental_health/suicide- prevention/world_report_2014/en/ (accessed 17 Decem- ber 2014).

5.HarrisEC,BarracloughB. Excess mortality of mental disorder. Br J Psychiatry 1998;173:11–53.

6.Isomets€ aET,HenrikssonMM,AroHM,HeikkinenME, KuoppasalmiKI,L€ onnqvistJK. Suicide in major depres- sion. Am J Psychiatry 1994;151:530–536.

7.TondoL,LepriB,BaldessariniRJ. Suicidal risks among 2826 Sardinian major aﬀective disorder patients. Acta Psychiatr Scand 2007;116:419–428.

8.NovickDM,SwartzH,FrankE. Suicide attempts in bipolar I and bipolar II disorder: a review and meta-ana- lysis of the evidence. Bipolar Disord 2010;12:1–9.

9.SchafferA,Isomets€ aET,TondoL et al. International Society for Bipolar Disorders Task Force on Suicide:

meta-analyses and meta-regression of correlates of sui- cide attempts and suicide deaths in bipolar disorder. Bipolar Disord 2015;17:1–16. 182 Tondo et al. 10.TondoL,BaldessariniRJ. Suicide in bipolar disorder. In:

YildizA,RuizP,NemeroffC, eds The bipolar book:

history, neurobiology, and treatment. New York, NY:

Oxford University Press; 2015: pp 509–530.

11.KoslowSH,RuizP,NemeroffCB, eds Concise guide to understanding suicide: epidemiology, pathophysiology, and prevention. New York, NY: Cambridge University Press; 2014.

12.YildizA,RuizP,NemeroffC, eds The bipolar book: his- tory, neurobiology, and treatment. New York, NY:

Oxford University Press; 2015.

13. daSilva CostaL,AlencarAP,NetoPJN et al. Risk fac- tors for suicide in bipolar disorder: systematic review. J Aﬀect Disord 2015;170:237–254.

14.RihmerA,GondaX,BalazsJ,FaludiG. The importance of depressive mixed states in suicidal behaviour. Neu- ropsychopharmacol Hung 2008;10:45–49.

15.Nordstr€ omP,AsbergM,Aberg-WistedtA,NordinC.

Attempted suicide predicts suicide risk in mood disor- ders. Acta Psychiatr Scand 1995;92:345–350.

16.HawtonK,SuttonL,HawC,SinclairJ,HarrissL. Sui- cide and attempted suicide in bipolar disorder: a system- atic review of risk factors. J Clin Psychiatry 2005;66:693–704.

17.BergenH,HawtonK,WatersK et al. Premature death after self-harm: multicentre cohort study. Lancet 2012;380 :1568–1574.

18.KesslerRC,BerglundP,BorgesG,NockM,WangPS.

Trends in suicide ideation, plans, gestures, and attempts in the United States, 1990–1992 to 2001–2003. JAMA 2005;293:2487–2495.

19.BaldessariniRJ,TondoL,DavisP,PompiliM,Goodwin FK,HennenJ. Decreased risk of suicides and attempts during long-term lithium treatment: a meta-analytic review. Bipolar Disord 2006;8:625–639.

20.GoodwinFK,JamisonKR. Manic-depressive illness:

bipolar disorders and recurrent depression. Oxford, NY:

Oxford University Press, 2007.

21. American Psychiatric Association. Practice guideline for the assessment and treatment of patients with suicidal behaviors. Am J Psychiatry 2003;160:1–60.

22.OquendoMA,EllisSP,GreenwaldS,MaloneKM, WeissmanMM,MannJJ. Ethnic and sex diﬀerences in suicide rates relative to major depression in the United States. Am J Psychiatry 2001;158:1652–1658.

23.BaldessariniRJ,VietaE,CalabreseJR,TohenM,Bow- denCL. Bipolar depression: overview and commentary.

Harv Rev Psychiatry 2010;18:143–157.

24.SaundersKEA,HawtonK. Clinical assessment and crisis intervention for the suicidal bipolar disorder patient.

Bipolar Disord 2013;15:575–583. 25.JuddLL,AkiskalHS,SchettlerPJ et al. The long-term natural history of the weekly symptomatic status of bipolar I disorder. Arch Gen Psychiatry 2002;59:530– 537.

26.ForteA,BaldessariniRJ,TondoL,V azquezGH,Pompili M,GirardiP. Long-term morbidity in bipolar-I, bipolar- II, and unipolar major depressive disorders. J Aﬀect Dis- ord 2015;178:71–78.

27.KupkaR,AltshulerL,NolenW. Three times more days depressed than manic or hypomanic in both bipolar I and bipolar II disorder1. Bipolar Disord 2007;9:531– 535.

28.BaldessariniRJ,SalvatoreP,KhalsaH-MK,TohenM.

Dissimilar morbidity following initial mania versus mixed-states in type-I bipolar disorder. J Aﬀect Disord 2010;126:299–302.29.De DiosC,EzquiagaE,GarciaA,SolerB,VietaE. Time spent with symptoms in a cohort of bipolar disorder out- patients in Spain: a prospective, 18-month follow-up study. J Aﬀect Disord 2010;125:74–81.

30.€ OsbyU,BrandtL,CorreiaN,EkbomA,Spar enP. Excess mortality in bipolar and unipolar disorder in Sweden.

Arch Gen Psychiatry 2001;58:844–850.

31.HayesJF,Miles J,WaltersK,KingM,OsbornDPJ. Sys- tematic review and meta-analysis of premature mortality in bipolar aﬀective disorder. Acta Psychiatr Scand 2015;131:417–425.

32.DunnerDL,GershonES,GoodwinFK. Heritable factors in the severity of aﬀective illness. Biol Psychiatry 1976;11:31–42.

33.HanusK,Zaplet alekM. Suicidal activity of patients with aﬀective disorders during the preventive use of lithium (Czech). Cesk Psychiatr 1984;80:97–100.

34.CoryellW,AndreasenNC,EndicottJ,KellerM. Signif- icance of past mania or hypomania in the course and outcome of major depression. Am J Psychiatry 1987;144:309–315.

35.Roy-ByrnePP,PostRM,HambrickDD,LeverichGS, RosoffAS. Suicide and course of illness in major aﬀec- tive disorder. J Aﬀect Disord 1988;15:1–8.

36.WehrTA,SackDA,RosenthalNE,CowdryRW. Rapid cycling aﬀective disorder: contributing factors and treat- ment responses in 51 patients. Am J Psychiatry 1988;145:179–184.

37.RihmerZ,Sz ant oK,BarsiJ. Suicide prevention: fact or ﬁction. Br J Psychiatry 1993;162:130– 131.

38.WuH,DunnerDL. Suicide attempts in rapid cycling bipolar disorder patients. J Aﬀect Disord 1993;29:

57–61.

39.ChenYW,DilsaverSC. Lifetime rates of suicide attempts among subjects with bipolar and unipolar dis- orders relative to subjects with other Axis I disorders.

Biol Psychiatry 1996;39:896–899.

40.FeinmanJA,DunnerDL. The eﬀect of alcohol and sub- stance abuse on the course of bipolar aﬀective disorder.

J Aﬀect Disord 1996;37:43–49.

41.BocchettaA,ArdauR,BurraiC,ChillottiC,Quesada G,Del ZompoM. Suicidal behavior on and oﬀ lithium prophylaxis in a group of patients with prior suicide attempts. J Clin Psychopharmacol 1998;18:384– 389.

42.PerugiG,AkiskalHS,MicheliC et al. Clinical subtypes of bipolar mixed states: validating a broader European deﬁnition in 143 cases. J Aﬀect Disord 1997;43:169–180.

43.VietaE,BenabarreA,ColomF et al. Suicidal behavior in bipolar I and bipolar II disorder. J Nerv Ment Dis 1997;185:407–409.

44.GoldbergJF,GarnoJL,LeonAC,KocsisJH,PorteraL.

Association of recurrent suicidal ideation with nonremis- sion from acute mixed mania. Am J Psychiatry 1998;155:1753–1755.

45.CassanoGB,Dell’OssoL, FrankE et al. The bipolar spectrum: a clinical reality in search of diagnostic criteria and an assessment methodology. J Aﬀect Disord 1999;54:319–328.

46.TondoL,BaldessariniRJ,HennenJ et al. Suicide attempts in major aﬀective disorder patients with comor- bid substance use disorders. J Clin Psychiatry 1999;60 (Suppl 2):63–69.

47.TsaiSY,LeeJC,ChenCC. Characteristics and psychoso- cial problems of patients with bipolar disorder at high risk for suicide attempt. J Aﬀect Disord 1999;52:145– 152. 183 Suicide attempts in bipolar disorders 48.PerugiG,MicheliC,AkiskalHS et al. Polarity of the ﬁrst episode, clinical characteristics, and course of manic depressive illness: systematic retrospective investigation of 320 bipolar I patients. Compr Psychiatry 2000;41:13– 18.

49.HenryC,SorbaraF,LacosteJ,GindreC,LeboyerM.

Antidepressant-induced mania in bipolar patients: iden- tiﬁcation of risk factors. J Clin Psychiatry 2001;62:249– 255.

50.SuppesT,LeverichGS,KeckPE et al. The Stanley Foun- dation Bipolar Treatment Outcome Network. II. demo- graphics and illness characteristics of the ﬁrst 261 patients. J Aﬀect Disord 2001;67:45–59.

51.VietaE,ColomF,CorbellaB et al. Clinical correlates of psychiatric comorbidity in bipolar I patients. Bipolar Disord 2001;3:253–258.

52.DittmannS,BiedermannNC,GrunzeH et al. The Stanley Foundation Bipolar Network: results of a naturalistic follow-up study after 2.5 years of follow-up in the German centres. Neuropsychobiology 2002;46(Suppl 1):2–9.

53.RucciP,FrankE,KostelnikB et al. Suicide attempts in patients with bipolar I disorder during acute and mainte- nance phases of intensive treatment with pharmacother- apy and adjunctive psychotherapy. Am J Psychiatry 2002;159:1160–1164.

54.SerrettiA,MandelliL,LattuadaE,CusinC,SmeraldiE.

Clinical and demographic features of mood disorder subtypes. Psychiatry Res 2002;112:195–210.

55.Ten Have M,VolleberghW,BijlR,NolenWA. Bipolar disorder in the general population in the Netherlands (prevalence, consequences and care utilisation): results from the Netherlands Mental Health Survey and Inci- dence Study (NEMESIS). J Aﬀect Disord 2002;68:203– 213.

56.BowdenCL,CalabreseJR,SachsG et al. A placebo-con- trolled 18-month trial of lamotrigine and lithium mainte- nance treatment in recently manic or hypomanic patients with bipolar I disorder. Arch Gen Psychiatry 2003;60:392–400.

57.CalabreseJR,BowdenCL,SachsG et al. A placebo-con- trolled 18-month trial of lamotrigine and lithium mainte- nancetreatment in recently manic or hypomanic patients with bipolar I disorder. Arch Gen Psychiatry 2003;61:680.

58.Cate CarterTD,MundoE,ParikhSV,KennedyJL. Early age at onset as a risk factor for poor outcome of bipolar disorder. J Psychiatr Res 2003;37:297–303.

59.GoodwinFK,FiremanB,SimonGE,HunkelerEM,LeeJ, RevickiD. Suicide risk in bipolar disorder during treat- ment with lithium and divalproex. JAMA 2003;290:1467–1473.

60.HenryC,Van den BulkeD,BellivierF,EtainB,Rouil- lonF,LeboyerM. Anxiety disorders in 318 bipolar patients: prevalence and impact on illness severity and response to mood stabilizer. J Clin Psychiatry 2003;64:331–335.

61.JuddLL,AkiskalHS. The prevalence and disability of bipolar spectrum disorders in the US population: re-ana- lysis of the ECA database taking into account subthresh- old cases. J Aﬀect Disord 2003;73:123– 131.

62.LeverichGS,AltshulerLL,FryeMA et al. Factors associated with suicide attempts in 648 patients with bipolar disorder in the Stanley Foundation Bipolar Net- work. J Clin Psychiatry 2003;64:506–515.

63.MacKinnonDF,ZandiPP,GershonE,NurnbergerJI, ReichT,DePauloJR. Rapid switching of mood in fami-lies with multiple cases of bipolar disorder. Arch Gen Psychiatry 2003;60:921–928.

64.SakamotoK,FukunagaT. The impact of attempted sui- cide on the symptoms and course of mood disorders. J Clin Psychiatry 2003;64:1210–1216.

65.YerevanianBI,KoekRJ,MintzJ. Lithium, anticonvul- sants and suicidal behavior in bipolar disorder. J Aﬀect Disord 2003;73:223–228.

66.DaltonEJ,Cate-CarterTD,MundoE,ParikhSV,Ken- nedyJL. Suicide risk in bipolar patients: the role of co- morbid substance use disorders. Bipolar Disord 2003;5:58–61.

67.Engstr€ omC,Br€ andstr€ omS,SigvardssonS,CloningerCR, NylanderPO. Bipolar disorder. III: harm avoidance a risk factor for suicide attempts. Bipolar Disord 2004;6:130–138.

68.ErnstCL,GoldbergJF. Clinical features related to age at onset in bipolar disorder. J Aﬀect Disord 2004; 82:21– 27.

69.FagioliniA,KupferDJ,RucciP,ScottJA,NovickDM, FrankE. Suicide attempts and ideation in patients with bipolar I disorder. J Clin Psychiatry 2004;65:509–514.

70.JoycePR,LutySE,McKenzieJM et al. Bipolar II disor- der: personality and outcome in two clinical samples.

Aust N Z J Psychiatry 2004;38:433–438.

71.PerlisRH,MiyaharaS,MarangellLB et al. Long-term implications of early onset in bipolar disorder: data from the ﬁrst 1000 participants in the Systematic Treatment Enhancement Program for Bipolar disorder (STEP-BD).

Biol Psychiatry 2004;55:875–881.

72.RajaM,AzzoniDA. Suicide attempts: diﬀerences between unipolar and bipolar patients and among groups with diﬀerent lethality risk. J Aﬀect Disord 2004;82:437–442.

73.RanM-S,WuQ-H,ConwellY,ChenEY-H,ChanCL- W. Suicidal behavior among inpatients with schizophre- nia and mood disorders in Chengdu, China. Suicide Life Threat Behav 2004;34:311–319.

74.ShiL,ThiebaudP,McCombsJS. The impact of unrecog- nized bipolar disorders for patients treated for depres- sion with antidepressants in the fee-for-services California Medicaid (Medi-Cal) program. J Aﬀect Dis- ord 2004;82:373–383.

75.SlamaF,BellivierF,HenryC et al. Bipolar patients with suicidal behavior: toward the identiﬁcation of a clinical subgroup. J Clin Psychiatry 2004;65:1035–1039. 76.VietaE,CalabreseJR,HennenJ et al. Comparison of rapid-cycling and non-rapid-cycling bipolar I manic patients during treatment with olanzapine: analysis of pooled data. J Clin Psychiatry 2004;65:1420–1428.

77.AngstJ,AngstF,Gerber-WerderR,GammaA. Suicide in 406 mood-disorder patients with and without long- term medication: a 40 to 44 years’ follow-up. Arch Sui- cide Res 2005;9:279–300.

78.HajekT,SlaneyC,GarnhamJ,RuzickovaM,Passmore M,AldaM. Clinical correlates of current level of func- tioning in primary care-treated bipolar patients. Bipolar Disord 2005;7:286–291.

79.KawaI,CarterJD,JoycePR et al. Gender diﬀerences in bipolar disorder: age of onset, course, comorbidity, and symptom presentation. Bipolar Disord 2005;7:119–125.

80.KupkaRW,LuckenbaughDA,PostRM et al. Compar- ison of rapid-cycling and non-rapid-cycling bipolar dis- order based on prospective mood ratings in 539 outpatients. Am J Psychiatry 2005;162:1273–1280.

81.MorenoDH,AndradeLH. Lifetime prevalence, health services utilization and risk of suicide of bipolar spec- 184 Tondo et al. trum subjects, including subthreshold categories in the Sao Paulo ECA study. J Aﬀect Disord 2005;87:231–241.

82.Valenc aAM,NardiAE,NascimentoI et al. Do social anxiety disorder patients belong to a bipolar spectrum subgroup? J Aﬀect Disord 2005;86:11–18.

83.WeissRD,OstacherMJ,OttoMW et al. Does recovery from substance use disorder matter in patients with bipolar disorder? J Clin Psychiatry 2005;66:730–735.

84.GalfalvyH,OquendoMA,CarballoJJ et al. Clinical predictors of suicidal acts after major depression in bipo- lar disorder: prospective study. Bipolar Disord 2006;8:586–595.

85.Gonzalez-PintoA,MosqueraF,AlonsoM et al. Suicidal risk in bipolar I disorder patients and adherence to long- term lithium treatment. Bipolar Disord 2006;8:618–624.

86.GrunebaumMF,RamsaySR,GalfalvyHC et al. Corre- lates of suicide attempt history in bipolar disorder: a stress-diathesis perspective. Bipolar Disord 2006;8:551– 557.

87.ValtonenHM,SuominenK,MantereO,Lepp€ am€ akiS, ArvilommiP,Isomets€ aET. Prospective study of risk fac- tors for attempted suicide among patients with bipolar disorder. Bipolar Disord 2006;8:576–585.

88.BaderCD,DunnerDL. Bipolar disorder not otherwise speciﬁed in relation to the bipolar spectrum. Bipolar Disord 2007;9 :860–867.

89.GazalleFK,HallalPC,TramontinaJ et al. Polyphar- macy and suicide attempts in bipolar disorder. Rev Bras Psiquiatr 2007;29:35–38.

90.LopezV,Detera-WadleighS,CardonaI,KassemL, McMahonFJ. Nested association between genetic varia- tion in tryptophan hydroxylase II: bipolar aﬀective dis- order, and suicide attempts. Biol Psychiatry 2007;61:181–186.

91.SimonGE,HunkelerE,FiremanB,LeeJY,SavarinoJ.

Risk of suicide attempt and suicide death in patients treated for bipolar disorder. Bipolar Disord 2007;9:526– 530.

92.SimonNM,ZaltaAK,OttoMW et al. The association of comorbid anxiety disorders with suicide attempts and suicidal ideation in outpatients with bipolar disorder. J Psychiatr Res 2007;41:255–264.

93.AzorinJ-M,KaladjianA,AdidaM et al. Factors associ- ated with rapid cycling in bipolar I manic patients: ﬁnd- ings from a French national study. CNS Spectr 2008;13:780–787.

94.CarballoJJ,Harkavy-FriedmanJ,BurkeAK et al. Fam- ily history of suicidal behavior and early traumatic expe- riences: additive eﬀect on suicidality and course of bipolar illness? J Aﬀect Disord 2008;109:57–63.

95.CruzN,VietaE,ComesM,HaroJM,ReedC,BertschJ.

Rapid-cycling bipolar I disorder: course and treatment outcome of a large sample across Europe. J Psychiatr Res 2008; 42:1068–1075.

96.GaoK,VerduinML,KempDE et al. Clinical correlates of patients with rapid-cycling bipolar disorder and a recent history of substance use disorder: a subtype comparison from baseline data of 2 randomized, pla- cebo-controlled trials. J Clin Psychiatry 2008;69:1057– 1063.

97.NakagawaA,GrunebaumMF,SullivanGM et al.

Comorbid anxiety in bipolar disorder: does it have an independent eﬀect on suicidality? Bipolar Disord 2008;10:530–538.

98.PompiliM,InnamoratiM,MannJJ et al. Periventricular white matter hyperintensities as predictors of suicide attempts in bipolar disorders and unipolar depression.Prog Neuropsychopharmacology Biol Psychiatry 2008;32:1501–1507.

99.RajaM,AzzoniA,KoukopoulosAE. Psychopharmaco- logical treatment before suicide attempt among patients admitted to a psychiatric intensive care unit. J Aﬀect Disord 2008;113:37–44.

100.SajatovicM,BiswasK,KilbourneAK,FennH,Williford W,BauerMS. Factors associated with prospective long- term treatment adherence among individuals with bipo- lar disorder. Psychiatr Serv 2008;59:753–759.

101.SubletteEM,CarballoJJ,MorenoC et al. Substance use disorders and suicide attempts in bipolar subtypes. J Psychiatr Res 2008;43:230–238.

102.AzorinJ-M,KaladjianA,AdidaM et al. Risk factors associated with lifetime suicide attempts in bipolar I patients: ﬁndings from a French national cohort. Compr Psychiatry 2009;50:115 –120.

103.Malloy-DinizLF,NevesFS,AbrantesSSC,FuentesD, Corr^ eaH. Suicide behavior and neuropsychological assessment of type I bipolar patients. J Aﬀect Disord 2009;112:231–236.

104.RajaM,AzzoniA,KoukopoulosAE. Psychopharmaco- logical treatment before suicide attempt among patients admitted to a psychiatric intensive care unit. J Aﬀect Disord 2009;113:37–44.

105.S anchez-GistauV,ColomF,Man eA,RomeroS,Sugra- nyesG,VietaE. Atypical depression is associated with suicide attempt in bipolar disorder. Acta Psychiatr Scand 2009;120:30–36.

106.MonteleoneP,BifulcoM,MainaG et al. Investigation of CNR1 and FAAH endocannabinoid gene polymor- phisms in bipolar disorder and major depression. Phar- macol Res 2010;61:400–404.

107.RyuV,JonD-I,ChoHS et al. Initial depressive episodes aﬀect the risk of suicide attempts in Korean patients with bipolar disorder. Yonsei Med J 2010;51:641–647.

108.BellivierF,YonL,LuquiensA et al. Suicidal attempts in bipolar disorder: results from an observational study (EMBLEM). Bipolar Disord 2011;13:377–386.

109.DennehyEB, MarangellLB,AllenMH,ChessickC, WisniewskiSR,ThaseME. Suicide and suicide attempts in the Systematic Treatment Enhancement Program for Bipolar Disorder (STEP-BD). J Aﬀect Disord 2011;133:423–427.

110.MitchellPB,FranklandA,Hadzi-PavlovicD et al.

Comparison of depressive episodes in bipolar disorder and in major depressive disorder within bipolar disorder pedigrees. Br J Psychiatry 2011;199:303–309.

111.PerroudN,BaudP,MouthonD,CourtetP,Malafosse A. Impulsivity, aggression and suicidal behavior in unipolar and bipolar disorders. J Aﬀect Disord 2011;134:112–118.

112.BegaS,SchafferA,GoldsteinB,LevittA. Diﬀerentiat- ing between bipolar disorder types I and II: results from the National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). J Aﬀect Disord 2012;138:46–53.

113.FinsethPI,MorkenG,AndreassenOA,MaltUF,Vaaler AE. Risk factors related to lifetime suicide attempts in acutely admitted bipolar disorder inpatients. Bipolar Disord 2012;14:727–734.

114.ParmentierC,EtainB,YonL, et al. Clinical and dimen- sional characteristics of euthymic bipolar patients with or without suicidal behavior. Eur Psychiatry 2012;27:570–576.

115.AntypaN,AntonioliM,SerrettiA. Clinical, psychologi- cal and environmental predictors of prospective suicide 185 Suicide attempts in bipolar disorders events in patients with bipolar disorder. J Psychiatr Res 2013;47:1800–1808.

116.Bani-FatemiA,PolsinelliG,KennedyJL,De LucaV.

Ethnicity and suicide attempt: analysis in bipolar disor- der and schizophrenia. BMC Psychiatry 2013;13:252.

117.ShabaniA,TeimurinejadS,KokarS et al. Suicide risk fac- tors in Iranian patients with bipolar disorder: 21-month follow-up study. Iran J Psychiatry Behav Sci 2013;7:16– 23.

118.BenedettiF,RiccaboniR,PolettiS et al. Serotonin transporter genotype modulates the relationship between early stress and adult suicidality in bipolar dis- order. Bipolar Disord 2014;16:857–866.

119.HolmaKM,HaukkaJ,SuominenK et al. Diﬀerences in incidence of suicide attempts between bipolar I and II disorders and major depressive disorder. Bipolar Disord 2014;16:652–661.

120.KurumajiA,NarushimaK,OoshimaK,YukizaneT, TakedaM,NishikawaT. Clinical course of the bipolar II disorder in a Japanese sample. J Aﬀect Disord 2014;168:363–366.

121.NeryFG,Miranda-ScippaA,Nery-FernandesF,Kapczin- skiF,LaferB. Prevalence and clinical correlates of alco- hol use disorders among bipolar disorder patients:

results from the Brazilian Bipolar Research Network.

Compr Psychiatry 2014;55:1116–1121.

122.SandbergJV,JakobssonJ,P alssonE,Land enM,Math e AA. Low neuropeptide Y in cerebrospinal ﬂuid in bipo- lar patients is associated with previous and prospective suicide attempts. Eur Neuropsychopharmacol 2014;24:

1–9.

123.ShibreT,HanlonC,MedhinG et al. Suicide and suicide attempts in people with severe mental disorders in Buta- jira, Ethiopia: 10 year follow-up of a population-based cohort. BMC Psychiatry 2014;14:150.

124.ZimmermanM,MartinezJ,YoungD,ChelminskiI,Mor- ganTA,DalrympleK. Comorbid bipolar disorder and borderline personality disorder and history of suicide attempts. J Pers Disord 2014;28:358–364.

125.Da Grac a CantarelliM,NardinP,BuffonA et al.

Serum triglycerides, but not cholesterol or leptin, are decreased in suicide attempters with mood disorders. J Aﬀect Disord 2015;172:403–409.

126.G€ ucl€ uO,S enormanc € O,Ayd nE,Erk ranM,K€ okt€ urk F. Phenomenological subtypes of mania and their rela-tionships with substance use disorders. J Aﬀect Disord 2015;174:569–573.

127.KimJS,HaTH,ChangJS et al. Seasonality and its dis- tinct clinical correlates in bipolar II disorder. Psychiatry Res 2015;225:540 –544.

128.Oli eE,SeyllerM,BeziatS et al. Clinical and neuropsy- chological characteristics of euthymic bipolar patients having a history of severe suicide attempt. Acta Psychi- atr Scand 2014;131:129–138.

129.TundoA,CalabreseJR,ProiettiL, deFillippisR. Varia- tion in response to short-term antidepressant treatment between patients with continuous and non-continuous cycling bipolar disorders. J Aﬀect Disord 2015;174:126– 130.

130.ZengR,CohenLJ,TanisT et al. Assessing the contribu- tion of borderline personality disorder and features to suicide risk in psychiatric inpatients with bipolar disor- der, major depression and schizoaﬀective disorder. Psy- chiatry Res 2015;226:361–367.

131.NordentoftM,MortensenPB,PedersenCB. Absolute risk of suicide after ﬁrst hospital contact in mental disor- der. Arch Gen Psychiatry 2011;68:1058–1064.

132.TondoL,LepriB,CruzN,BaldessariniRJ. Age at onset in 3014 Sardinian bipolar and major depressive disor- der patients. Acta Psychiatr Scand 2010;121:446–452.

133.BaldessariniRJ,TondoL,VazquezGH et al. Age at onset versus family history and clinical outcomes in 1665 international bipolar-I disorder patients. World Psychiatry 2012;11:40–46.

134.CrosbyAE,HanB,OrtegaLAG,ParksSE,GfroererJ.

Suicidal thoughts and behaviors among adults≥18 years –United States, 2008–2009. MMWR Surveill Summ 2011;60:1–22, http://www.cdc.gov/mmwr/preview/ mmwrhtml/ss6013a1.htm.

135.BaethgeC,GruschkaP,SmolkaMN et al. Eﬀectiveness and outcome predictors of long-term lithium prophy- laxis in unipolar major depressive disorder. J Psychiatry Neurosci 2003;28:355–361.

136.BaethgeC,TondoL,BrattiIM et al. Prophylaxis latency and outcome in bipolar disorders. Can J Psychi- atry 2003;48:449–457.

137.BrattiIM,BaldessariniRJ,BaethgeC,TondoL. Pre- treatment episode count and response to lithium treat- ment in manic-depressive illness. Harv Rev Psychiatry 2003;11:245–256. 186 Tondo et al. Copyright ofActa Psychiatrica Scandinavica isthe property ofWiley- Blackwell andits content maynotbecopied oremailed tomultiple sitesorposted toalistserv without the copyright holder'sexpresswrittenpermission. However,usersmayprint, download, oremail articles forindividual use.