BRITISH JOURNAL OF (2006), 189, 515^515^519. 519. doi: 10.1192/bjp.bp.105.020321

Influence of sub-syndromal symptoms after adequate therapeutic treatment in the man- agement of these symptoms. remission from manic or mixed episodes A critical part of developing effective treatment strategies for the management of both sub-syndromal and syndromal MAURICIO TOHEN, CHARLES L. BOWDEN, JOSEPH R. CALABRESE, symptoms requires a greater understanding DANIEL LIN, TAMMY D. FORRESTER, GARY S. SACHS, of the factors involved in their develop- ATHANASIOS KOUKOPOULOS, LAKSHMI YATHAM and HEINZ GRUNZE ment. The goal of our post hocposthoc analysesanalyses was to identify factors that are associated with a greater percentage of time with sub-syndromal symptoms, and to charac- terise how these symptoms influence Background Sub-syndromal Recent attention has focused on the need to outcomes in a randomised double-blind symptoms in impair develop therapeutic strategies that allow clinical trial of relapse prevention compar- patients with bipolar affective disorder to ing and lithium in patients with functioning and diminish quality of life. return to a premorbid level of function. . Aims Toexamine factors associated Effective therapies are currently available for the treatment of acute manic and de- with time spent with sub-syndromal pressive episodes, and for prolonging time METHOD symptoms and to characterise how these in remission. However, patients considered symptoms influence outcomes. to have responded to treatment may The details of this randomised double-blind nevertheless continue to experience sub- controlled trial have been described pre- MethodMethod In a double-blind randomised syndromal symptoms that impair function- viously (Tohen et aletal, 2005) and are sum- maintenance trial, patients received either ing and diminish quality of life. This dispar- marised briefly here. Participants were at olanzapine orlithiumor lithium monotherapy for1 ity between symptom amelioration and least 18 years old and met DSM–IV functional outcomes among patients with (American Psychiatric Association, 1994) year. Stepwise logistic regression models bipolar disorder has been described by criteria for an index manic or mixed bi- were used to identify factors that were TohenTohen et aletal (1990(1990aa) and Chengappa et aletal polar I episode based on the Structured significant predictors of percentage time (2005). In a study of first-episode disorder, Clinical Interview for DSM–IV (First et aletal,, spent with sub-syndromal symptoms.The TohenTohen et aletal (2000) observed that only 38% 1995). Prerequisites for study entry were a presence of sub-syndromal symptoms of patients achieved functional recovery baseline total score of 20 or more on the within 2 years of treatment for an acute Young Rating Scale (YMRS; Young during the first 8 weekswas examined as a manic episode. One factor that may et aletal, 1978) and a history of at least two predictor of subsequent relapse. contribute to the difficulty in reaching full manic or mixed episodes within the functional recovery is the presence of sub- previous 6 years. ResultsResults Presence of sub-syndromal syndromal symptoms that either individu- The trial consisted of four study depressive symptoms during the first ally or as an aggregate are not sufficiently periods: screening (2–7 days); open-label 8 weeks significantlyincreased the severe to constitute a major mood episode, co-therapy (6–12 weeks); double-blind likelihood of depressive relapse (relative but interfere with functioning. taper (4 weeks); and double-blind mono- A recent reassessment of bipolar- therapy (48 weeks). The starting daily risk 4.67, PP550.001). Patients with spectrum disorders including patients with dosages for open-label co-therapy were psychotic features and those with a sub-syndromal symptoms revealed at least olanzapine 15 mg and lithium 600 mg. Sub- greatergreaternumberofpreviousdepressive number of previous depressive a five-fold greater prevalence than found sequent dosages of olanzapine could range episodes were more likely to experience with traditionally defined syndromal from 5 mg to 20 mg per day. Investigators sub-syndromal depressive symptoms diagnoses (Judd & Akiskal, 2003). The were required to optimise lithium dosage persistence of sub-syndromal symptoms in and reach a target serum level of 0.6– (RR(RR¼2.51,2.51, PP550.001and RR¼2.35,2.35, patients with bipolar disorder has been 1.2 mmol/l by week 4 during this period. PP¼0.03 respectively). shown to contribute substantially to func- Patients who met symptomatic remis- tional impairment (Gitlin et aletal, 1995;,1995; sion criteria for bipolar disorder – a total Conclusions These findings help to AltshulerAltshuler et aletal, 2002). Furthermore, sub- YMRS score of 12 or below and a total identify patients atincreasedat increased riskof syndromal symptoms are associated with score of 8 or below on the 21-item affective relapse and suggestthat an increased risk of relapse (Goodnick etet Hamilton Rating Scale for Depression appropriate therapeutic interventions alal, 1987; Tohen et aletal, 1990a; Keller et aletal,, (HRSD; Hamilton, 1959) – during the 1992). Keller et aletal (1992) demonstrated second study period (co-therapy) were should be considered even when that patients prescribed lithium who randomly assigned to treatment with either syndromal-level symptoms are absent. achieved standard serum levels of the drug olanzapine or lithium monotherapy. were both less likely to develop sub-syndro- During the third study period, patients Declaration of interest Sponsorship mal symptoms and less likely to experience remained on their current dosage of from Lilly Research Laboratories. relapse relative to those who achieved low- randomised treatment and the dosage of range levels, which highlights the role of the discontinued drug was tapered over

515515

Downloaded from https://www.cambridge.org/core. 27 Sep 2021 at 17:50:25, subject to the Cambridge Core terms of use. TOHEN EET T AL

4 weeks. During the final study period, Ta b l e 1 Categorical definitions of euthymia, sub-syndromal status and relapse lithium levels were monitored and the dosage adjusted to maintain serum levels EuthymiaSub-syndromal Relapse in the therapeutic range 0.6–1.2 mmol/l. All patients randomly assigned to olanza- ManiaYMRS 448Y8 YMRSMRS 9^149^14 YMRS 5515 pine also had blood drawn to maintain DepressionHRSD 448H8 HRSDRSD 9^149^14 HRSD 5515 the study masking. For every outlier report MixedMixedYMRS 448andHRSD448YMRS9^14andHRSD9^148 YMRS 9^14 and HRSD 9^14YMRS 5515 and HRSD 551515 generated for a patient in the lithium group, Bipolar YMRS 448andHRSD448YMRS 9^14 or HRSD 9^14YMRS 5515 or HRSD 5515 a sham lithium outlier report was sent to a disorderdisorder patient in the olanzapine group. Thus, re- ports to investigative sites indicating that HRSD, Hamilton Rating Scale for Depression (21-item); YMRS,Young Mania Rating Scale. the lithium dosage should be adjusted did not unmask the randomised assignment. staying in the model. Odds ratio estimates no statistically significant difference be- Illness severity was assessed using the and their corresponding 95% confidence tween therapies with respect to percentage YMRS and HRSD.The categorical defini- intervals were calculated for explanatory of patients with sub-syndromal symptoms tions of euthymia, sub-syndromal symp- variables in the final model. at any time, or the percentage of time with toms and relapse (Table 1) are based on The presence of sub-syndromalsub-syndromal symp- sub-syndromal bipolar symptoms overall or previous reports (Yatham et aletal, 2004;,2004; tomstomsduringduring the first 8 weeks was examined in the individual poles (depression or GopalGopal et aletal, 2005; Tohen et aletal, 2005).,2005). as a predictor of subsequent relapse. Sepa- mania). The mean percentages of time Relapse and sub-syndromal symptoms were rate analyses were made for each type of spent with any mood symptom or with classified globally as bipolar (any mood sub-syndromal status (any episode, depres- manic or depressive sub-syndromal symp- symptom) and by their respective manic sive, manic) and each type of relapse (any toms among patients who experienced and depressive poles. episode, depressive, manic). Patients who them were 27.4% (s.d.¼25.7), 26.8% completed at least 8 weeks of therapy with- (s.d.(s.d.¼28.7), and 24.4% (s.d.¼18.1) respec- out relapse (nn¼340) were included in the tively (median times 18.4%, 15.4% and Statistical methods analysis. Additional analyses evaluated the 21.5% respectively). Stepwise logistic regression models were potential impact of the presence of residual Among patients who completed the used to identify factors that were significant symptoms at the outset of the double-blind first 8 weeks of the relapse prevention predictors of the proportion of time spent treatment phase and of the absence of phase without a major affective episode with sub-syndromal symptoms. The symptoms at the initiation of the double- ((nn¼340), the subsequent rate of relapse response, percentage time spent with sub- blind periodfollowed by the emergence for those with bipolar sub-syndromal syndromal symptoms, was broken into five of sub-ofsub-syndromalsyndromal symptoms during the symptoms was 36.8% (32 of 87) and categories by quartiles: 0%, 0–5525%,25%, first 8 weeks on the subsequent risk of re- 29.6% (75 of 253) for those without. 25%–25%–5550%, 50%–5575% and 75% or lapse: patients were stratified according to Regardless of therapy, the presence of more. Linear logistic regression models for the presence (YMRS score 9–14) or absence sub-syndromal bipolar symptoms signifi- these ordinal response data were fitted by (YMRS score 448) of residual symptoms at cantly increased the likelihood of relapse the method of maximum likelihood using the outset of the double-blind period. Fish- into the depressive pole (17 of 87 patients SAS (version 8.2) PROC LOGISTIC. Cate- er’s exact test was used to test proportions, with sub-syndromal bipolar symptoms v.v. gorical explanatory variables included ther- and relapse incidence rates were charac- 26 of 253 patients without such symptoms; apy, gender, index episode type (mania or terised by an estimate of the relative risk relative risk 1.9, 95% CI 1.09–3.33, mixed),mixed),presencepresence of psychotic features, with 95% confidence limits; estimates were PP¼0.038). The presence of sub-syndromal rapid-rapid-cyclingcycling status, proneness to depres- constructed with patients having no time depressive (but not manic) symptoms also sion or mania (defined by the predominant with sub-syndromal symptoms during the increased the likelihood of bipolar relapse episode type in previous episodes), number first 8 weeks as the referent. (15 of 26 patients with symptoms v.v. 92 of92of of previous bipolar episodes (0–5, 6–9, 10 314 patients without; RR¼1.97, 95% CI or more), number of previous manic epi- RESULTSRESULTS 1.36–2.85,1.36–2.85, PP¼0.004) and relapse into the sodes (0–2, 3–5, 6 or more) and number depressive (but not manic) pole (12 of 26 of previous depressive episodes (0–1, 2–3, Our analyses included 424 patients with patients with symptoms v.v. 31 of 314 4 or more). Dimensional explanatory vari- bipolar I disorder. Their baseline demo- patients without; RR¼4.67, 95% CI ables included age, onset age, duration of graphic and illness characteristics are 2.74–7.97,2.74–7.97, PP550.001). Patients who started illness, baseline YMRS total score, and presented in Table 2. No statistically signif- the relapse prevention phase without sub- baseline HRSD total score. The variable icant difference was observed between the syndromal bipolar symptoms but devel- therapy was included in all models regard- olanzapine and lithium treatment groups oped them during the first 8 weeks were less of statistical significance. The other on these baseline measures. significantly more likely to experience bi- explanatory variables were entered into The percentages of patients with sub- polar relapse relative to those without the model if they were significant at the syndromal symptoms at any time during sub-syndromal symptoms during the same aa¼0.05 level and could be removed in a the 48-week study are shown in Table 3. period (20 of 39 patients v.v. 75 of 25325375of stepwise manner if the least significant Presented in Table 4 are the percentages patients; RR¼1.73, 95% CI 1.21–2.48, effect in the model at a particular step did of time spent with sub-syndromal symp- PP¼0.01). This was especially true of not meet this level of significance for toms categorised by quartiles. There was patients with sub-syndromal depressive

516

Downloaded from https://www.cambridge.org/core. 27 Sep 2021 at 17:50:25, subject to the Cambridge Core terms of use. SUB -SYNDROMAL SYMPTOMS OF BIPOLAR DISORDER

Ta b l e 22Tab Demographic and illness characteristics of the sample Ta b l e 33Tab Participants with sub-syndromal symptomsatanytimeduringthe48-weekstudysymptoms at any time during the 48-week study Lithium group Olanzapine group ((nn¼211) ((nn¼213) Participants, nn (%)

Gender, nn (%)(%) Lithium Olanzapine FemaleFemale 114 (54.0)(54.0)114 112 (52.6) groupgroup group Male 97 (46.0) 10110 (47.4)1(47.4) nn¼211211 nn¼213 Mania type, nn (%) Any mood symptom 83 (39.3)(39.3)83 80 (37.6)80(37.6) MixedMixed 12 (5.7) 14 (6.6) Depression 39 (18.5)35 (16.4) Mania 199 (94.3)(94.3)199 199 (93.4) Mania 56 (26.5)56(26.5) 48 (22.5)48(22.5) Psychotic features present, nn (%) 51 (24.2)(24.2)51 58 (27.2)58(27.2) Rapid-cycling present, nn (%) 7(3.3)7 (3.3) 6(2.8) Proneness category, nn (%) Depression 73 (34.6) 72 (33.8)(33.8)72 previous depressive episodes were 1.96 Mania 138 (65.4)(65.4)138 141 (66.2)(66.2)141 (95% CI 1.04–3.71) times more likely to Previous bipolar episodes, nn (%) experience sub-syndromal depressive symp- 1^5 73 (34.6) 83 (39.0)(39.0)83 toms relative to those with no or one 6^96^9 7171(33.6) (33.6) 73 (34.3) previous episode, and those with four or more were 2.35 (95% CI 1.21–4.54) times 10 or more 67 (31.8)(31.8)67 57 (26.8)(26.8)57 more likely to do so than those with no or Previous manic episodes, nn (%)(%) one episode (PP¼0.03).0.03). 1^2 66 (31.3) 57 (26.8)(26.8)57 3^5 70 (33.2)(33.2)70 92 (43.2)(43.2)92 66ormore or more 75 (35.5)(35.5)75 64 (30.0)64(30.0) DISCUSSION Previous depressive episodes, nn (%) 0^10^1 73 (34.6) 89 (41.8)89(41.8) Our analyses identified illness characteris- 2-3 73 (34.6)(34.6)73 71 (33.3)(33.3)71 tics that were associated with a greater 4ormore 6565(30.8) (30.8) 53 (24.9)(24.9)53 amount of time spent experiencing sub- Number of previous bipolar episodes syndromal symptoms in patients with bi- Median 767 6 polar I disorder, and examined the impact IQRIQR 868 6 of these symptoms on outcomes. There was no statistically significant difference Number of previous manic episodes between the olanzapine and lithium treat- Median 444 4 ment groups with regard to the percentage IQRIQR 545 4 of patients who experienced sub-syndromal Number of previous depressive episodes symptoms or in the percentage of time Median 222 2 spent with symptoms. Approximately IQRIQR 323 2 38% of patients experienced symptoms Age, years: mean (s.d.) 42.2 (12.1) 42.6 (13.1) that fell within the sub-syndromal range Ageatonset,years:mean(s.d.)Age at onset, years: mean (s.d.) 28.228.2(10.1) (10.1) 28.9 (11.5)(11.5)28.9 of severity, as defined by rating scales, at Duration of illness, years: mean (s.d.) 13.5 (9.9)(9.9)13.5 13.3 (10.5) any time during the maintenance phase of YMRS total score at visit 1: mean (s.d.) 28.5 (6.1) 27.9 (5.5)(5.5)27.9 this study. Among these patients, just over HRSD total score at baseline: mean (s.d.) 5.6 (4.2)(4.2)5.6 5.4 (4.3) a quarter (27%) of the time was spent with sub-syndromal symptoms. These results are HRSD, Hamilton Rating Scale for Depression (21-item); IQR, interquartile range; YMRS,Young Mania Rating Scale. consistent with previous reports from non-controlled studies documenting the prevalence of sub-syndromal symptoms in symptoms and subsequent relapse into the patients; RR¼0.84, 95% CI 0.50–1.43, patients with bipolar I disorder (Keitner etet depressive pole (12 of 25 patients v.v. 31 of31of PP¼0.60).0.60). alal, 1996; Judd et aletal, 2002; Post et aletal,, 314 patients; RR¼4.86, 95% CI 2.87– Of the individual factors assessed, 2003). Our findings further extend the view 8.24,8.24, PP550.0001). In contrast, patients presence of psychotic features and a greater that sub-syndromal symptoms are common who had residual sub-syndromal symptoms number of previous depressive episodes and pervasive in bipolar I disorder, even in (any mood) at the outset of the double- were associated with increased time with a population of patients who achieved blind phase and continued to experience depressive sub-syndromal symptoms. clinical response from an acute manic or symptoms during the first 8 weeks were Patients with psychotic features were 2.51 mixed episode and continued to receive not more likely to relapse than those with- (95% CI 1.47–4.30; PP550.001) times more treatment for relapse prevention. out sub-syndromal symptoms during this likely to experience sub-syndromal depres- Meaningful comparisons between these period (12 of 48 patients v.v. 75 of 25325375of sive symptoms. Patients with two or three findings and those of previous reports

517

Downloaded from https://www.cambridge.org/core. 27 Sep 2021 at 17:50:25, subject to the Cambridge Core terms of use. TOHEN EET T AL

Ta b l e 4 Percentage of time spent with sub-syndromal symptoms sub-syndromal manic symptoms and manic relapse. In contrast, in a study of relapse Time with symptoms, % prevention with lithium, Keller et aletal (1992) reported in a non-controlled study nn 0%0% 0^25%0^25% 25^50%25^50%50^75% 4475%75% a stronger association between manic sub- syndromal symptoms and manic relapse re- BipolarBipolar lative to the depressive polarity. It is not Lithium 21160.7 22.322.310.0 5.7 1.4 clear what factors account for these discre- Olanzapine 213213 62.462.4 22.122.17.0 3.8 4.7 pant results; however, it is possible that the Depression use of olanzapine and lithium in combina- Lithium 21181.5 12.3 4.3 1.9 0.0 tion to treat the acute episode, and mono- Olanzapine 21383.6 8.5 5.6 1.4 0.9 therapy during relapse prevention, might Mania have contributed to our results. Further Lithium 21173.5 16.1 4.7 4.7 0.9 analyses of sub-syndromal symptoms dur- ing the first 8 weeks of the monotherapy Olanzapine 21377.5 14.614.62.3 2.3 3.3 phase that differentiated residual sub- syndromal symptoms (i.e. symptoms from the index episode that had not resolved with respect to time with sub-syndromal without such features. This finding is inter- completely) from newly emerged sub- symptoms are difficult to make because of esting in light of recent reports questioning syndromal symptoms yielded intriguing differences in defining criteria, methods of the prognostic value of psychotic features in results: the emergence of sub-syndromal assessment and study design. Notably, for bipolar disorder with respect to illness symptoms (in particular sub-syndromal participants in our study who experienced severity and treatment response (Keck etet depressive symptoms) was associated with relapse into an affective episode during alal, 2003). In a study by Swann et aletal significantly greater risk of subsequent the study period, only data prior to the (2004) the presence of psychotic features relapse into the depressive pole, whereas event were included in the analyses. It was associated with greater overall func- the presence of residual symptoms was should also be noted that all patients en- tional impairment, but was not correlated not. This finding suggests that the longitu- tered this study with an index manic or with higher baseline mania scores or altered dinal assessment of symptom severity, as mixed episode, which may account in part treatment response. On the other hand, a opposed to just cross-sectional assessment, for differences in the distribution of time study by Tohen et aletal (1990(1990bb) identified)identified might better determine the risk of with symptoms in the two poles relative the presence of psychotic features during subsequent relapse. to studies that involved patients with either the index episode as a predictor of shorter manic/mixed or depressive episodes. In- time in remission. deed, previous studies have reported that A greater number of previous depres- Limitations patients spend substantially more time with sive episodes was also associated with a There are several limitations to these ana- depressive symptoms than with manic greater percentage of time spent with sub- lyses that warrant discussion. Patients in symptoms (Judd et aletal, 2002; Post et aletal,, syndromal depressive symptoms, which is this study were required to achieve remis- 2003; Joffe2003;Joffe et aletal, 2004), whereas the distri- in accordance with a previous report by sion from an acute manic or mixed episode bution of time with symptoms in our study PostPost et aletal (2003). These findings also extend to be included in the relapse prevention was roughly equal for the manic (26.8%) previous reports that a greater number of phase of the trial; thus, this population con- and depressive (24.4%) poles. Since the previous affective episodes increases the sisted of patients who responded to com- polarity of index mood episode is a predic- risk of subsequent relapse (Kessing et aletal,, bined olanzapine and lithium treatment tor of the polarity of subsequent relapse 2004).2004). and who might not be representative of (Tohen(Tohen et aletal, 2003), it is possible that the the general population of people with bi- distribution of sub-syndromal symptoms is polar I disorder. A related limitation is that similarly dependent on the polarity of the Predictors of relapse our results may not be generalised to pa- index episode. In our study the presence of sub-syndromalsyndromalsub- tients who are unable or refuse to partici- symptoms during the first 8 weeks of the pate in clinical trials. It should be noted, relapse prevention phase was associated however, that in contrast to naturalistic Predictors of sub-syndromal with a significantly greater likelihood of studies, the analyses of data from our symptoms subsequent relapse, particularly into the randomised clinical trial were adjusted Of the clinical variables analysed, the pre- depressive pole. This finding agrees, in part, for pharmacological treatment. Another sence of psychotic features and the with previous studies that have reported an limitation is the relatively short follow-up number of previous depressive episodes increased risk of relapse associated with period of only 1 year, which may not were associated with increased time spent sub-syndromal symptoms (Goodnick et aletal,, characterise the full longitudinal course of with sub-syndromal symptoms. Patients 1987; Tohen et aletal, 1990,1990aa; Keller;Keller et aletal,, bipolar disorder. who entered the study with psychotic 1992). However, the presence of depressive Our findings provide prognostic value features were more likely to experience sub-syndromal symptoms was predictive of in terms of identifying patients at increased a greater percentage of time with sub- depressive relapse, whereas there was risk of affective relapse. Given that the syndromal depressive symptoms than those no corresponding relationship between presence of sub-syndromal symptoms is

518

Downloaded from https://www.cambridge.org/core. 27 Sep 2021 at 17:50:25, subject to the Cambridge Core terms of use. SUB -SYNDROMAL SYMPTOMS OF BIPOLAR DISORDER

associated with significant functional im- MAURICIOTOHEN, MD,DrPH, Lilly Research Laboratories, Indianapolis, Indiana and Department of pairment (Tohen et aletal, 1990,1990aa; Altshuler;Altshuler etet Psychiatry,McLean Hospital, Harvard Medical School, Belmont, Massachusetts; CHARLES L. BOWDEN, MD, alal, 2002) and incurs large social and Department of Psychiatry,University of Texas Health Sciences Center, San Antonio,Texas; JOSEPH R. economic health costs (Bauer et aletal, 2001;,2001; CALABRESE, MD, Department of Psychiatry,Case Western Reserve University,Cleveland, Ohio; DANIEL LIN, MacQueenMacQueen et aletal, 2003), appropriate PhD,TAMMY D.FORRESTER, Lilly Research Laboratories, Indianapolis, Indiana; GARYS. SACHS, MD, pharmacological and non-pharmacological Department of Psychiatry,Harvard Medical School and Massachusetts General Hospital, Boston, therapeutic interventions should be consid- Massachusetts,USA;Massachusetts, USA; ATHANASIOS KOUKOPOULOS, MD,Centro Lucio Bini, Rome, Italy; LAKSHMI YATHAM, ered even in the absence of symptoms at the MD,University of British Columbia,Vancouver,Canada; HEINZGRUNZE, MD,Department of Psychiatry, University of Munich, Germany syndromal level. Correspondence: Dr MauricioTohen,Lilly Research Laboratories,Indianapolis,IN 46285,USA.Tel: REFERENCES +1(317) 277 9585; fax: +1(317) 276 7845; email: m.tohen@@lilly.comlilly.com

Altshuler, L. L., Gitlin, M. J., Mintz, J., et aletal (2002)(2002) (First received 2 December 2005, final revision 5 June 2006, accepted 4 July 2006) Subsyndromal depression is associated with functional impairment in patients with bipolar disorder. Journal ofofJournal Clinical Psychiatry,, 6363,,807^811. 807^811. Judd,Judd,L.L.&Akiskal,H.S. L. L. & Akiskal, H. S. (2003)(2003) The prevalence and Swann, A. C., Daniel, D. G., Kochan, L. D., et aletal American Psychiatric Association(19 (1994) 94) Diagnostic disability of bipolar spectrum disorders in the US (2004)(2004) Psychosis in mania: specificity of its role in and Statistical Manual of Mental Disorders (4th edn) population: re-analysis of the ECA database taking into severity and treatment response. Journal of Clinical (DSM^IV).Washington, DC: APA. account subthreshold cases. Journal of Affective Disorders,, PsychiatryPsychiatry,, 65,,825^829. 825^829. 7373,123^131. Bauer, M. S., Kirk, G. F.,Gavin,F., Gavin, C., et aletal (2001)(2001) Tohen, M.,Waternaux, C. M. & Tsuang, M.M.T. T.(19 (1990 9 0 aa)) Determinants of functional outcome and healthcare Judd,Judd,L.L.,Akiskal,H.S.,Schettler,P.J., L. L., Akiskal, H. S., Schettler, P.J., et aletal (2002)(2002) Outcome in mania. A 4-year prospective follow-up of costs in bipolar disorder: a high-intensity follow-up The long-term natural history of the weekly 75 patients utilizing survival analysis. Archives of General study.study. Journal of Affective Disorders,, 6565,231^241., 231^241. symptomatic status of bipolar I disorder. Archives ofofArchives PsychiatryPsychiatry,, 47, 110 6 ^ 1111.1111.,110 General Psychiatry,, 5959,530^537. Chengappa, K. N., Hennen, J., Baldessarini, R. J., etet Tohen, M.,Waternaux, C. M.,Tsuang,M., Tsuang, M. T., et aletal al (2005)(2005) Recovery and functional outcomes following Keck,P.E.,McElroy,S.L.,Havens,J.R.,et aletal (2003)(2003) (19 9 0 bb)) Four-year follow-up of twenty-four first- olanzapine treatment for bipolar I mania. Bipolar Psychosis in bipolar disorder: phenomenology and episode manic patients. Journal of Affective Disorders,, 1919,, DisordersDisorders,, 77,,68^76. 68^76. impact on morbidity and course of illness. 79^86.79^86. First, M. B., Spitzer, R. L., Gibbon, M., et aletal (19 95) Comprehensive Psychiatry,, 44, 263^269. Tohen, M., Hennen, J., Zarate, C. M., et aletal (2000) Structured Clinical Interview for DSM ^ IV Axis Disorders ^ Keitner,G. I., Solomon, D. A., Ryan,C. E., et aletal (19 9 6) Two-year syndromal and functional recovery in 219 Patient Edition. New York: New York State Psychiatric Prodromal and residual symptoms in bipolar I disorder. cases of first-episode major affective disorder with Institute.Institute. Comprehensive Psychiatry,, 3737, 362^367.,362^367. psychotic features. American Journal of Psychiatry,, 157,, Gitlin, M. J., Swendsen, J., Heller,T. L., et aletal (19 95) 220^228.220^228. Relapse and impairment in bipolar disorder. AmericanAmerican Keller,M.B.,Lavori,P.W.,Kane,J.M.,Keller, M. B., Lavori, P.W., Kane, J. M., et aletal (19 92)9 2)(19 Journal of Psychiatry,, 152152,1635^1640. Subsyndromal symptoms in bipolar disorder. A Tohen, M., Zarate,C. A., Hennen, J., et aletal (2003)(2003) TheThe comparison of standard and low serum levels of lithium. McLean^Harvard First-Episode Mania Study: Goodnick, P.J., Fieve, R. R., Schlegel, A., et aletal (19 87) Archives of General Psychiatry,, 49, 371^376.,371^376. prediction of recovery and first recurrence. AmericanAmerican Predictors of interepisode symptoms and relapse in Journal of Psychiatry,, 160160, 2099^2107. affective disorder patients treated with lithium Kessing, L.V., Hansen, M. G., Andersen, P.K., et aletal carbonate.carbonate. American Journal of Psychiatry,, 144, 367^369.,367^369. (2004)(2004) The predictive effect of episodes on the risk of Tohen, M., Greil,W., Calabrese, J. R., et al (2005)(2005) recurrence in depressive and bipolar disorders ^ a life- Olanzapine versus lithium in the maintenance treatment Gopal, S., Steffens, D. C., Kramer, M. L., et aletal (2005)(2005) long perspective. Acta Psychiatrica Scandinavica,, 109109,, of bipolar disorder: a 12-month, randomized, double- Symptomatic remission in patients with bipolar mania: 339^344.339^344. blind, controlled clinical trial. American Journal of results from a double-blind, placebo-controlled trial of PsychiatryPsychiatry,, 162162,1281^1290. risperidone monotherapy. Journal of Clinical Psychiatry,, MacQueen, G. M., Marriott, M., Begin, H., et aletal 6666,1016^1020. (2003)(2003) Subsyndromal symptoms assessed in longitudinal, Yatham, L. N., Binder,C., Kusumakar,V., et al (2004)(2004) Hamilton, M.(1959) The assessment of anxiety states prospective follow-up of a cohort of patients with Risperidone plus lithium versus risperidone plus by rating.byrating. British Journal of Medical Psychology,, 32, 50^55.,50^55. bipolar disorder. Bipolar Disorders,, 55, 349^355.,349^355. valproate in acute and continuation treatment of mania. InternationalClinicalPsychopharmacologyInternational Clinical ,, 1919,103^109. Joffe,Joffe,R.T.,MacQueen,G.M.,Marriott,M., R.T., MacQueen, G. M., Marriott, M., et aletal Post, R. M., Denicoff, K. D., Leverich, G. S., et aletal (2004)(2004) A prospective, longitudinal study of percentage (2003)(2003) Morbidity in 258 bipolar outpatients followed for Young, R. C., Biggs, J.T., Ziegler,V.E., et aletal (1978)(19 78) AA of time spent ill in patients with bipolar I or bipolar II 1 year with daily prospective ratings on the NIMH life rating scale for mania: reliability, validity and sensitivity. disorders. Bipolar Disorders,, 66,,62^66. 62^66. chart method. Journal of Clinical Psychiatry,, 6464,680^690., 680^690. British Journal of Psychiatry,, 133133, 429^435.,429^435.

519519

Downloaded from https://www.cambridge.org/core. 27 Sep 2021 at 17:50:25, subject to the Cambridge Core terms of use.