Course and outcome of BPAD - NIMHANS copy.pptx

Bpad course

1. Course and outcome of bipolar affective disorder Ravi Philip Rajkumar Associate Professor, Department of Psychiatry, JIPMER, Puducherry Time to relapse Cycle length Rapid cycling % of time spent in illness Predominant polarity Age at onset

2. Outline History Definitions and other terms Research methods Prodrome and short-term course ( ≤ 2 years) Long-term course Outcomes Predictors of course and outcome Future directions

3. An ancient debate... Is bipolar disorder.... “a form of mental illness with an episodic course, full recovery, and no deterioration in personality or lasting deficits?” (Emil Kraepelin) “...a terrible, incurable form of insanity? ” (Jules Falret ) OR

4. First, some definitions Course is the longitudinal description of the illness Outcome is the clinical or functional end-point at a specified period after the onset of illness; therefore, it is cross-sectional

5. Terms used to describe the course of bipolar disorder Age at onset Number of episodes Episode frequency ( ≈ cycle length) Spontaneous cycling Time to mania / depression / any relapse Rapid cycling (and its variants) Seasonality There are certainly others!

6. Types of outcome Clinical or symptomatic recovered, syndromal, subsyndromal? Functional social, occupational, marital / familial... Neurocognitive deficits and their progression... Mortality any cause, medical, accidental, suicide... Morbidity (?) higher risks of certain medical disorders...

7. Remember.... Clinical outcome (control of symptoms) DOES NOT ALWAYS CORRELATE WITH Functional outcome (employment, marriage, day-to-day functioning, quality of life)

8. How do we study course? Longitudinal follow-up studies best, but also most resource-intensive... Retrospective and catch-up studies easier, but subject to bias and missing data... Chart reviews easiest, but of relatively lower value

9. Studies of short-term course and outcome

10. The bipolar prodrome Subsyndromal or non-specific symptoms In children and adolescents (Egeland et al., 2000) In adults (Zeschel et al., 2013) Episodic sad mood Episodic irritability Anger dyscontrol Episodic changes in energy level Demanding behaviour Conduct symptoms Sleep disturbance In mania Increased energy level Racing thoughts Agitation Overtalkativeness In depression Low mood Reduced energy Fatigue Insomnia

11. First-episode mania: time to recovery – United States (Tohen et al., AJP 2003)

12. First-episode mania: time to relapse – United States (Tohen et al., AJP 2003)

13. 1-year outcomes following admission – United States (Keck et al., AJP 1998)

14. 2-year outcomes for bipolar I disorder: Australia (Kulkarni et al., BMC Psychiatry 2012)

15. 2-year outcomes for bipolar I disorder: EMBLEM study (France) (Azorin et al., BMC Psychiatry 2009)

16. Cognition: short-term outcomes 29 stable patients with BP, 30 healthy controls 3-month follow-up Baseline deficits in attention , processing speed HAM-D, YMRS, cognition assessed Over follow-up : Change in depressive symptoms associated with decreased verbal fluency No cognitive effect of changes in manic symptoms (Chaves et al., 2011)

17. Cognition: short-term outcomes 53 patients with first episode mania 1 year follow-up Deficits in verbal, nonverbal, working memory at baseline Patients who remained well over 1 year showed improvements in cognition Recurrences were associated with further declines (Kozicky et al., 2014)

18. Functioning: short-term outcomes Study Study sample and follow-up Results Goetz et al., 2007 (EMBLEM) 3681 patients with acute mania; 1 year 28% good work outcome; 68% high work impairment Reed et al., 2010 (EMBLEM) 1393 patients with acute mania; 2 years 41% high work impairment; good outcome correlated with independent housing, living together Bearden et al., 2011 79 recovered BP-I; 9 months 45 (57%) occupationally recovered at baseline; a further 8 (10%) recovered at follow-up

19. Studies of long-term course and outcome

20. What happens to “soft” bipolar disorders? 57 patients with cyclothymia or BPAD-NOS 4.5 years follow-up 42.1% changed to bipolar type II 10.5% changed to bipolar type I (Alloy et al., 2012)

21. What happens to “soft” bipolar disorders? 140 youth with BPAD-NOS 5 years follow-up 23% changed to bipolar type I 22% changed to bipolar type II (Axelson et al., 2011)

22. How stable is the diagnosis of bipolar disorder? Over 10 years, 77.8% of 95 patients with BP retained their diagnosis at final follow-up However, this was unchanged only in 69.5%; 8.4% had diagnostic disputes in the interval. In a study of first-episode BP patients over 10 years, only 50.3% maintained the diagnosis at all follow-ups; 49.7% had at least one change in between. (Bromet et al., 2011; Ruggero et al., 2010)

23. How much time do patients spend in different phases? Bipolar I disorder 146 patients Follow-up 12.8 years Symptomatic 47.3% of the time 31.9% depressive, 8.9% manic, 5.9% mixed Subsyndromal symptoms 3 times more common than full syndromes Bipolar II disorder 86 patients Follow-up 13.4 years Symptomatic 53.9% of the time 50.3% depressive, 1.3% hypomania, 2.3% mixed Subsyndromal symptoms 3 times more common than full syndromes (Judd et al., 2002, 2003)

24. Do the symptom patterns change with age? Depressive symptoms are more prominent in the third, fourth and fifth decades of life. Time spent in mania is not influenced by age. (Coryell et al., 2009)

25. How long do episodes last? 219 patients with bipolar I disorder; 5-year follow-up * Only two cases (Solomon et al., 2010) Episode type Time to recovery for 50% patients (weeks) Time to recovery for 75% patients (weeks) Mania 7 15 Depression 15 35 Hypomania 3 6 Mixed* 7 76

26. Do cycle lengths shorten with time? A review of older reports (summarized in Baldessarini et al., 2012) found evidence of shortening in approximately 1/3 of 40 published studies. Study Study sample and follow-up Results Turvey et al., 1999 165 patients with BP-I; 10 years No shortening observed Cusin et al., 2000 244 patients with BP, 182 with depression; 14 years Shortening in both groups, but more prominent in BP Dittmann et al., 2002 152 patients with BP, 2.5 years Shortening in the majority of patients Baldessarini et al., 2012 128 patients with BP, 5.7 years Random course common; shortening rare Subramaniam et al. (unpublished) 150 patients with BP; retrospective Weak evidence of shortening

27. Course on treatment: 20 years ago (United States) 4.3 year follow-up of 82 outpatients with DSM-III bipolar disorder 73% relapse rate 37% relapsed by 1 year 2/3 of these (49%) had multiple relapses Mean survival time 2.92 years Relapses not due to drug default (Gitlin et al., 1995)

28. Course on treatment: the latest results (Austria) 4-year follow-up of 300 patients (158 BP-I, 142 BP-II) 68% relapse rate Mean time to relapse: 208 days Not much difference between BP-I and BP-II Lithium prolongs time to any relapse Either patient- or doctor-initiated medication changes shorten the time to relapse (!) (Simhandl et al., JCP 2014)

29. Course of rapid-cycling bipolar disorder 89 patients with rapid cycling BP, 13.7 year follow-up More weeks in depression / hypomania Higher likelihood of serious suicidal attempts In 80% of cases, rapid cycling resolved over 2 years (Coryell et al., 2003)

30. Course of cognitive deficits 68 BP-I, 45 healthy controls ≈9 year follow-up Baseline impaired attention, memory, executive functions Attention improved slightly over time Executive function worsened with time (Torrent et al., 2012)

31. Course of cognitive deficits: older adults 33 patients with BP, mean age 69.7 years 36 healthy controls 3-year follow-up Accelerated cognitive decline compared to controls (Gildengers et al., 2009)

32. But are we sure? Meta-analysis of studies from 1990-2014 Mean follow-up period 4.62 years 14 cognitive measures Scores remained stable Accept the null hypothesis? (Samame et al., 2014)

33. Functioning: long-term outcomes Study Study sample and follow-up period Result Coryell et al., 1993 148 BP, 240 depression; 5 years Significant fall in job status, income even after recovery; BP=MDD Goldberg et al., 1995 51 BP, 49 depression; 4.5 years Good outcome in only 41% of BP; significant work disability Judd et al., 2005 158 BP-I, 133 BP-II; 15 years Symptom improvement and functioning related; depression, not mania, predicts disability Goldberg and Harrow 2010 34 BP, 89 depression; 10 years Good outcome in 50% of BP; functional decline in 30-40% Burdick et al., 2010 33 BP, 15 years Poor work outcome in 15 (45.5%) patients Goldberg and Harrow 2011 46 BP, 49 depression; 15 years Good outcome in 35% BP vs 73% depression

34. How common is “poor-outcome” bipolar disorder? Earlier results suggest that around 15-20% of patients have poor functional outcomes (Hastings, 1958; Tsuang et al., 1979) A later meta-analysis found that at least 30% of patients do not reach premorbid levels (MacQueen et al., 2001) A recent study of out-patients found that 14 of 54 (26%) had severe disability (Rosa et al., 2014)

35. What happens to antidepressant-induced switches? Study Sample description Results Navarro et al., 2014 2-year follow-up of patients who developed hypomania on antidepressants No new cases of hypomania, despite antidepressant continuation Wada et al., 2013 1-year follow-up of 33 patients admitted for antidepressant-induced hypomania or mania 6 patients (18.2%) converted to bipolar disorder

36. Seasonality Seasonal peak Reported in Summer peak of mania United Kingdom, Australia, Taiwan Spring peak of mania United States Late winter-spring peak of mania Brazil, India* Winter peak of mania North India No specific pattern for mania India, England, Canada, and one multi-centric study spanning both hemispheres Winter peak of depression England No specific pattern for depression Canada Summer peak for depression North India Potential influence for many variables: day length, sunlight exposure, humidity, rainfall, latitude... * Unpublished data

37. Suicide as an outcome Suicide attempts in 182 of 4360 patients Completed suicide in 8 of these (Dennehy et al., STEP-BD, 2011) 8 deaths from suicide in a UK cohort of 235 patients, collected over a 35-year period (Dutta et al., 2007)

38. Predictors of suicide Early onset Family history of suicide Depressive symptoms / mixed states Severe episodes Rapid cycling Axis I comorbidity Substance use comorbidity (Hawton et al., 2005)

39. Violent crime as an outcome Systematic review: odds ratio of 2-9 for a patient with BP to commit a violent crime Prospective study: 8% (314 of 3743) Swedish patients with BP committed violent crime. However, this was significant only in those with comorbid substance use. No relationship to episode type / symptom profile (Fazel et al., 2010)

40. Death as an outcome 10922 Taiwanese patients with either depression or bipolar disorder (1542 BP) 1.3 fold increase in overall mortality in BP 4-fold increased risk of suicide 2-fold increased risk of accidental death Cardiovascular comorbidity correlates with suicide (Chang et al., 2012)

41. What is the scenario in India? Chopra et al., 2006 Retrospective life chart study of 34 patients Community health setting (Sakalawara) Mean 4.4 episodes over 20.2 years 72% of episodes were mania Mean duration of mania: 13 weeks Mean duration of depression: 18 weeks Cycle length shortened, but not significantly

42. What is the scenario in India? Subramanian et al., 2015 Retrospective life chart study of 150 patients General hospital psychiatry unit (JIPMER) Mean 5.7 episodes over 13.4 years Mania ≈ 5 times commoner than depression 52.7% had only recurrent mania Median time to relapse: 24 months Weak trend to shortening of cycle length in time

43. Predictors of course and outcome

44. Age at onset Higher episode frequency Shorter time to relapse More time spent in depression More suicidal attempts Poorer functioning and quality of life More rapid cycling Higher substance use comorbidity Onset < 13 years has the greatest impact (Kupka et al., 2005; Perlis et al., 2009; Lev-Ran et al., 2013)

45. Gender Men Women More alcohol and drug use disorders More violent suicide attempts Longer time to recovery from acute mania Depressive onset Predominant depressive polarity More psychotic depression More rapid cycling More Axis II comorbidity More lifetime suicide attempts Faster recovery from acute mania Higher risk of depression after acute mania (Kupka et al., 2005; Miquel et al., 2011; Nivoli et al., 2011) ♂ ♀

46. Adverse life events 1. Childhood adversity Overall childhood trauma associated with early onset, poor functioning Physical and sexual abuse associated with a higher number of episodes Physical abuse also associated with self-harm and poor functioning Childhood physical / sexual abuse associated with rapid cycling (Kupka et al., 2005; Larsson et al., 2013)

47. Adverse life events 2. Current stressors Negative life events can trigger both unipolar and bipolar depression; not much difference In some studies, negative life events exacerbate depression Chronic stressors associated with depression, but not mania Low social support predicts depression No consistent evidence for an effect of life events on the onset or severity of mania (Weinstock and Miller, 2010; Gershon et al., 2013; reviewed in Johnson, 2005)

48. Comorbid substance use 1. Alcohol Alcohol use commonly precedes or accompanies bipolar depression Time sequence is important; “bipolar first” has a worse outcome than “alcohol first” Greater time spent in affective episodes Possible greater risk of cycling into mania Poorer response to lithium Greater risk of suicide (O'Connell et al., 1991; Strakowski et al., 2005; Baethge et al., 2008; Ostacher et al., 2010; Farren et al., 2011)

49. Comorbid substance use 2. Other substances Cannabis use commonly precedes or accompanies mania Possible greater risk of cycling into mania Higher rates of nicotine and alcohol use Higher rates of dissocial personality disorder Earlier onset Greater episode frequency and rapid cycling Greater risk of violent crime (Ostacher et al., 2010; Fazel et al., 2010; Lev-Ran et al., 2013)

50. Comorbid substance use 3. Effects of recovery In a sample of 1000 patients from STEP-BD, recovery from substance use was associated with improved role functioning. Treatment of comorbid BP and alcoholism with valproate may reduce heavy drinking and the number of drinks consumed over 6 months. (Weiss et al., 2005; Salloum et al., 2005)

51. Comorbid anxiety Three-fold increase in depressive morbidity Prolonged recovery from depression Risk of earlier relapse Poorer short-term (12 month) quality of life and role function Poorer 3-year global and illness outcomes Require more intensive psychotherapy during depressive episodes Worse impact for GAD and social phobia (Otto et al., 2006; Coryell et al., 2009; Deckersbach et al., 2014)

52. Family history of mood disorders Data from 2600 STEP-BD patients Earlier age at onset More rapid cycling More suicide attempts Elevated depressive symptoms Poorer quality of life Higher neuroticism and personality disorder traits (Antypa and Serretti, 2014)

53. Spontaneous cycling Switching polarity within an episode is associated with a poor overall outcome Polyphasic switching (more than 2 in an episode) associated with more time in an episode, and poor symptomatic and functional outcomes over 10 years. Some evidence of an association with substance use (Turvey et al., 1999; Maj et al., 2002; Vieta et al., 2009)

54. Psychotic symptoms Over a 4-year follow-up, 73.2% of patients with psychotic BP reached premorbid levels Baseline first-rank symptoms predicted a poorer outcome (Carlson et al., 2012) Mood-incongruent psychotic symptoms predict poorer symptomatic and functional outcomes (Miklowitz, 1992; Marneros et al., 2009)

55. Polarity of the index episode Retrospective study of 320 BP-I patients Index polarity appears to predict long-term polarity Rapid cycling and suicide attempts more common with depressive onset Psychotic symptoms less common with depressive onset Mixed onsets associated with suicide attempts and chronicity (Perugi et al., 2000)

56. Cognitive deficits Memory, scanning and processing speed predict short-term (9 month) work outcomes. Baseline processing speed predicts 15-year global outcome Verbal learning deficits predict long-term work outcome Baseline verbal memory deficits predict low functioning over 9 years. (Burdick et al., 2010; Bearden et al., 2011; Torrent et al., 2012)

57. Future directions

58. Staging bipolar disorder? Stage Description Stage I Return to premorbid status between episodes Stage II Have comorbidites or residual symptoms needing treatment; can manage daily activities Stage III Require social and occupational rehabilitation; have difficulty with daily activities Stage IV Patient unable to live independently or maintain self-care These stages correlate with cognitive impairment and structured assessments of functioning. (Kapczinski et al., 2009; Rosa et al., 2014)

59. Using technology to assess course? 62 patients with BP (47 BP-I, 15 BP-II), 36 weeks Text messages (n=54) or e-mail prompts (n=8) Weekly self-report; 75% compliant Results: 47.7% time in depression 7% in mania, 8.8% in mixed states 36.5% in euthymia Good agreement with existing results (Bopp et al., 2010)

60. Cross-national and cross-cultural differences? Indian data has shown a preponderance of mania, including “unipolar” mania, in contrast to “Western” studies Outcomes after a first episode of mania were uniformly better in Taiwan than in the United States (Strakowski et al., 2007; Narayanaswamy et al., 2014; Subramanian et al., 2015)

61. Tackling unanswered questions... Is there a bipolar “DOSMD challenge”? The “juvenile bipolar” controversy Genetic or biological markers of course Better delineation of the prodrome Tracking cognitive and real-world outcomes Novel predictors of outcome (sleep, body mass index)

62. And one final thought... Rapport et al., Acad. Psychiatry, 2013 121 charts of BP outpatients 41 treated by residents, 80 by consultants Improvement similar across groups Remission was 26/80 for consultants but only 5/41 for residents ....but???