Early  Interven,on  in  Bipolar  Disorder  

Ma5hew  Taylor  London  

Atlas  of  Indices  of  Depriva,on  2010  h5p://neighbourhood.sta,s,cs.gov.uk/  

Outline  

•  Why  do  we  need  to  intervene  earlier  in  bipolar  disorder?  – Diagnos,c  delays  – How  does  bipolar  disorder  fit  within  specialist  ‘Early  Interven,on’  services?  

•  What  are  the  future  direc,ons?  

Why  intervene  early  in  bipolar  disorder?  

•  Bipolar  disorder  is  a  major  problem  

•  Harm  from  recurrent  episodes  

•  Avoidable  delays  in  current  care  

Global  Burden  of  Disease:  Years  Lived  with  Disability  

1990   2010  

Articles

www.thelancet.com Vol 380 December 15/22/29, 2012 2163

Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: a systematic analysis for the Global Burden of Disease Study 2010

Lancet 2012; 380: 2163–96

See Comment pages 2053, 2054, 2055, 2058, 2060, 2062, and 2063

See Special Report page 2067

See Articles pages 2071, 2095, 2129, 2144, 2197, and 2224

*Authors listed alphabetically

†Joint senior authors

‡Corresponding author

SummaryBackground Non-fatal health outcomes from diseases and injuries are a crucial consideration in the promotion and monitoring of individual and population health. The Global Burden of Disease (GBD) studies done in 1990 and 2000 have been the only studies to quantify non-fatal health outcomes across an exhaustive set of disorders at the global and regional level. Neither eff ort quantifi ed uncertainty in prevalence or years lived with disability (YLDs).

Methods Of the 291 diseases and injuries in the GBD cause list, 289 cause disability. For 1160 sequelae of the 289 diseases and injuries, we undertook a systematic analysis of prevalence, incidence, remission, duration, and excess mortality. Sources included published studies, case notifi cation, population-based cancer registries, other disease registries, antenatal clinic serosurveillance, hospital discharge data, ambulatory care data, household surveys, other surveys, and cohort studies. For most sequelae, we used a Bayesian meta-regression method, DisMod-MR,

Theo Vos, Abraham D Flaxman, Mohsen Naghavi, Rafael Lozano, Catherine Michaud, Majid Ezzati, Kenji Shibuya, Joshua A Salomon, Safa Abdalla*, Victor Aboyans*, Jerry Abraham*, Ilana Ackerman*, Rakesh Aggarwal*, Stephanie Y Ahn*, Mohammed K Ali*, Miriam Alvarado*, H Ross Anderson*, Laurie M Anderson*, Kathryn G Andrews*, Charles Atkinson*, Larry M Baddour*, Adil N Bahalim*, Suzanne Barker-Collo*, Lope H Barrero*, David H Bartels*, Maria-Gloria Basáñez*, Amanda Baxter*, Michelle L Bell*, Emelia J Benjamin*, Derrick Bennett*, Eduardo Bernabé*, Kavi Bhalla*, Bishal Bhandari*, Boris Bikbov*, Aref Bin Abdulhak*, Gretchen Birbeck*, James A Black*, Hannah Blencowe*, Jed D Blore*, Fiona Blyth*, Ian Bolliger*, Audrey Bonaventure*, Soufi ane Boufous*, Rupert Bourne*, Michel Boussinesq*, Tasanee Braithwaite*, Carol Brayne*, Lisa Bridgett*, Simon Brooker*, Peter Brooks*, Traolach S Brugha*, Claire Bryan-Hancock*, Chiara Bucello*, Rachelle Buchbinder*, Geo! rey Buckle*, Christine M Budke*, Michael Burch*, Peter Burney*, Roy Burstein*, Bianca Calabria*, Benjamin Campbell*, Charles E Canter*, Hélène Carabin*, Jonathan Carapetis*, Loreto Carmona*, Claudia Cella*, Fiona Charlson*, Honglei Chen*, Andrew Tai-Ann Cheng*, David Chou*, Sumeet S Chugh*, Luc E Co! eng*, Steven D Colan*, Samantha Colquhoun*, K Ellicott Colson*, John Condon*, Myles D Connor*, Leslie T Cooper*, Matthew Corriere*, Monica Cortinovis*, Karen Courville de Vaccaro*, William Couser*, Benjamin C Cowie*, Michael H Criqui*, Marita Cross*, Kaustubh C Dabhadkar*, Manu Dahiya*, Nabila Dahodwala*, James Damsere-Derry*, Goodarz Danaei*, Adrian Davis*, Diego De Leo*, Louisa Degenhardt*, Robert Dellavalle*, Allyne Delossantos*, Julie Denenberg*, Sarah Derrett*, Don C Des Jarlais*, Samath D Dharmaratne*, Mukesh Dherani*, Cesar Diaz-Torne*, Helen Dolk*, E Ray Dorsey*, Tim Driscoll*, Herbert Duber*, Beth Ebel*, Karen Edmond*, Alexis Elbaz*, Suad Eltahir Ali*, Holly Erskine*, Patricia J Erwin*, Patricia Espindola*, Stalin E Ewoigbokhan*, Farshad Farzadfar*, Valery Feigin*, David T Felson*, Alize Ferrari*, Cleusa P Ferri*, Eric M Fèvre*, Mariel M Finucane*, Seth Flaxman*, Louise Flood*, Kyle Foreman*, Mohammad H Forouzanfar*, Francis Gerry R Fowkes*, Richard Franklin*, Marlene Fransen*, Michael K Freeman*, Belinda J Gabbe*, Sherine E Gabriel*, Emmanuela Gakidou*, Hammad A Ganatra*, Bianca Garcia*, Flavio Gaspari*, Richard F Gillum*, Gerhard Gmel*, Richard Gosselin*, Rebecca Grainger*, Justina Groeger*, Francis Guillemin*, David Gunnell*, Ramyani Gupta*, Juanita Haagsma*, Holly Hagan*, Yara A Halasa*, Wayne Hall*, Diana Haring*, Josep Maria Haro*, James E Harrison*, Rasmus Havmoeller*, Roderick J Hay*, Hideki Higashi*, Catherine Hill*, Bruno Hoen*, Howard Ho! man*, Peter J Hotez*, Damian Hoy*, John J Huang*, Sydney E Ibeanusi*, Kathryn H Jacobsen*, Spencer L James*, Deborah Jarvis*, Rashmi Jasrasaria*, Sudha Jayaraman*, Nicole Johns*, Jost B Jonas*, Ganesan Karthikeyan*, Nicholas Kassebaum*, Norito Kawakami*, Andre Keren*, Jon-Paul Khoo*, Charles H King*, Lisa Marie Knowlton*, Olive Kobusingye*, Adofo Koranteng*, Rita Krishnamurthi*, Ratilal Lalloo*, Laura L Laslett*, Tim Lathlean*, Janet L Leasher*, Yong Yi Lee*, James Leigh*, Stephen S Lim*, Elizabeth Limb*, John Kent Lin*, Michael Lipnick*, Steven E Lipshultz*, Wei Liu*, Maria Loane*, Summer Lockett Ohno*, Ronan Lyons*, Jixiang Ma*, Jacqueline Mabweijano*, Michael F MacIntyre*, Reza Malekzadeh*, Leslie Mallinger*, Sivabalan Manivannan*, Wagner Marcenes*, Lyn March*, David J Margolis*, Guy B Marks*, Robin Marks*, Akira Matsumori*, Richard Matzopoulos*, Bongani M Mayosi*, John H McAnulty*, Mary M McDermott*, Neil McGill*, John McGrath*, Maria Elena Medina-Mora*, Michele Meltzer*, George A Mensah*, Tony R Merriman*, Ana-Claire Meyer*, Valeria Miglioli*, Matthew Miller*, Ted R Miller*, Philip B Mitchell*, Ana Olga Mocumbi*, Terrie E Mo" tt*, Ali A Mokdad*, Lorenzo Monasta*, Marcella Montico*, Maziar Moradi-Lakeh*, Andrew Moran*, Lidia Morawska*, Rintaro Mori*, Michele E Murdoch*, Michael K Mwaniki*, Kovin Naidoo*, M Nathan Nair*, Luigi Naldi*, K M Venkat Narayan*, Paul K Nelson*, Robert G Nelson*, Michael C Nevitt*, Charles R Newton*, Sandra Nolte*, Paul Norman*, Rosana Norman*, Martin O’Donnell*, Simon O’Hanlon*, Casey Olives*, Saad B Omer*, Katrina Ortblad*, Richard Osborne*, Doruk Ozgediz*, Andrew Page*, Bishnu Pahari*, Jeyaraj Durai Pandian*, Andrea Panozo Rivero*, Scott B Patten*, Neil Pearce*, Rogelio Perez Padilla*, Fernando Perez-Ruiz*, Norberto Perico*, Konrad Pesudovs*, David Phillips*, Michael R Phillips*, Kelsey Pierce*, Sébastien Pion*, Guilherme V Polanczyk*, Suzanne Polinder*, C Arden Pope III*, Svetlana Popova*, Esteban Porrini*, Farshad Pourmalek*, Martin Prince*, Rachel L Pullan*, Kapa D Ramaiah*, Dharani Ranganathan*, Homie Razavi*, Mathilda Regan*, Jürgen T Rehm*, David B Rein*, Guiseppe Remuzzi*, Kathryn Richardson*, Frederick P Rivara*, Thomas Roberts*, Carolyn Robinson*, Felipe Rodriguez De Leòn*, Luca Ronfani*, Robin Room*, Lisa C Rosenfeld*, Lesley Rushton*, Ralph L Sacco*, Sukanta Saha*, Uchechukwu Sampson*, Lidia Sanchez-Riera*, Ella Sanman*, David C Schwebel*, James Graham Scott*, Maria Segui-Gomez*, Saeid Shahraz*, Donald S Shepard*, Hwashin Shin*, Rupak Shivakoti*, David Singh*, Gitanjali M Singh*, Jasvinder A Singh*, Jessica Singleton*, David A Sleet*, Karen Sliwa*, Emma Smith*, Jennifer L Smith*, Nicolas J C Stapelberg*, Andrew Steer*, Timothy Steiner*, Wilma A Stolk*, Lars Jacob Stovner*, Christopher Sudfeld*, Sana Syed*, Giorgio Tamburlini*, Mohammad Tavakkoli*, Hugh R Taylor*, Jennifer A Taylor*, William J Taylor*, Bernadette Thomas*, W Murray Thomson*, George D Thurston*, Imad M Tleyjeh*, Marcello Tonelli*, Je! rey A Towbin*, Thomas Truelsen*, Miltiadis K Tsilimbaris*, Clotilde Ubeda*, Eduardo A Undurraga*, Marieke J van der Werf*, Jim van Os*, Monica S Vavilala*, N Venketasubramanian*, Mengru Wang*, Wenzhi Wang*, Kerrianne Watt*, David J Weatherall*, Martin A Weinstock*, Robert Weintraub*, Marc G Weisskopf*, Myrna M Weissman*, Richard A White*, Harvey Whiteford*, Steven T Wiersma*, James D Wilkinson*, Hywel C Williams*, Sean R M Williams*, Emma Witt*, Frederick Wolfe*, Anthony D Woolf*, Sarah Wulf*, Pon-Hsiu Yeh*, Anita K M Zaidi*, Zhi-Jie Zheng*, David Zonies*, Alan D Lopez†, Christopher J L Murray†‡

Life Expectancy at Birth for People with Serious MentalIllness and Other Major Disorders from a SecondaryMental Health Care Case Register in LondonChin-Kuo Chang1*, Richard D. Hayes1, Gayan Perera1, Mathew T. M. Broadbent2, Andrea C. Fernandes1,

William E. Lee3, Mathew Hotopf3, Robert Stewart1

1 King’s College London (Institute of Psychiatry), London, United Kingdom, 2 South London and Maudsley NHS Foundation Trust, London, United Kingdom, 3 King’s

College London, Academic Dept Psychological Medicine, Institute of Psychiatry, London, United Kingdom

Abstract

Objective: Despite improving healthcare, the gap in mortality between people with serious mental illness (SMI) and generalpopulation persists, especially for younger age groups. The electronic database from a large and comprehensive secondarymental healthcare provider in London was utilized to assess the impact of SMI diagnoses on life expectancy at birth.

Method: People who were diagnosed with SMI (schizophrenia, schizoaffective disorder, bipolar disorder), substance usedisorder, and depressive episode/disorder before the end of 2009 and under active review by the South London andMaudsley NHS Foundation Trust (SLAM) in southeast London during 2007–09 comprised the sample, retrieved by the SLAMCase Register Interactive Search (CRIS) system. We estimated life expectancy at birth for people with SMI and eachdiagnosis, from national mortality returns between 2007–09, using a life table method.

Results: A total of 31,719 eligible people, aged 15 years or older, with SMI were analyzed. Among them, 1,370 died during2007–09. Compared to national figures, all disorders were associated with substantially lower life expectancy: 8.0 to 14.6 lifeyears lost for men and 9.8 to 17.5 life years lost for women. Highest reductions were found for men with schizophrenia (14.6years lost) and women with schizoaffective disorders (17.5 years lost).

Conclusion: The impact of serious mental illness on life expectancy is marked and generally higher than similarly calculatedimpacts of well-recognised adverse exposures such as smoking, diabetes and obesity. Strategies to identify and preventcauses of premature death are urgently required.

Citation: Chang C-K, Hayes RD, Perera G, Broadbent MTM, Fernandes AC, et al. (2011) Life Expectancy at Birth for People with Serious Mental Illness and OtherMajor Disorders from a Secondary Mental Health Care Case Register in London. PLoS ONE 6(5): e19590. doi:10.1371/journal.pone.0019590

Editor: James G. Scott, The University of Queensland, Australia

Received January 24, 2011; Accepted April 1, 2011; Published May 18, 2011

Copyright: ! 2011 Chang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: The development of the SLAM BRC Case Register was funded by two Capital Awards from the UK National Institute for Health Research and is furthersupported through the BRC Nucleus funded jointly by the Guy’s and St Thomas’ Trustees and South London and Maudsley Special Trustees. C-KC, RH, AF, MB, MHand RS are funded by the National Institute for Health Research (NIHR) Specialist Biomedical Research Centre for Mental Health at the South London andMaudsley NHS Foundation Trust and Institute of Psychiatry, King’s College London. WL is funded by the UK Medical Research Council. The funders had no role instudy design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

* E-mail: chin-kuo.chang@kcl.ac.uk

Introduction

Concerns about premature mortality among people with mentaldisorders have been increasing [1–6]. Higher general and specificcauses of mortality in all or specific age groups have been identifiedfor people with serious mental illnesses (SMI, which might includeschizophrenia, schizoaffective disorder, bipolar disorder, anddepressive psychosis) in a range of settings [1,2,7–11], althoughmost report relative risks or rates of mortality rather than lifeexpectancy – which is arguably a more important index for shapingpolicy since it takes premature mortality into account and has beenwidely used for other risk exposures such as smoking, diabetes,obesity, ethnicity and socioeconomic status [12–17]. Of the fewstudies to measure this outcome, clients with SMI from publicmental health agencies in eight US states from 1997 to 2000 werefound to have lower life expectancies of 13 to more than 30 years

compared to the general population (depending on the year andstate) [2], and a loss of 8.8 life years (14.1 years for men and 5.7years for women) was estimated in a comparison between peopletreated for SMI and the general population in Massachusetts, US[1]. A Swedish study using a nationwide hospital discharge registryreported substantial gaps in life expectancy at age 30 for mainmental disorder categories compared to the general population,particularly for functional psychosis other than schizophrenia /affective psychosis (15.9 years lost), substance abuse (15.6 years lost),and organic psychosis (14.8 years lost) among men and organicpsychosis (22.6 years lost), mental retardation (14.7 years lost), andsubstance abuse (18.8 years lost) among women [18]. Preliminarystrategies for preventing premature deaths among people with SMIhave been suggested, emphasizing the management of suicide riskand physical illness, minimum polypharmacy, and improvement ofaccessibility to physical health care [19].

PLoS ONE | www.plosone.org 1 May 2011 | Volume 6 | Issue 5 | e19590

Life  expectancy  with  bipolar  disorder:    Males:    67.3      =  10  year  reduc,on  

 (95%  CI  66.1  to  68.5)    Females:    70.4      =  11  year  reduc,on  

 (95%  CI  69.5  to  71.4)  

Life Expectancy at Birth for People with Serious MentalIllness and Other Major Disorders from a SecondaryMental Health Care Case Register in LondonChin-Kuo Chang1*, Richard D. Hayes1, Gayan Perera1, Mathew T. M. Broadbent2, Andrea C. Fernandes1,

William E. Lee3, Mathew Hotopf3, Robert Stewart1

1 King’s College London (Institute of Psychiatry), London, United Kingdom, 2 South London and Maudsley NHS Foundation Trust, London, United Kingdom, 3 King’s

College London, Academic Dept Psychological Medicine, Institute of Psychiatry, London, United Kingdom

Abstract

Objective: Despite improving healthcare, the gap in mortality between people with serious mental illness (SMI) and generalpopulation persists, especially for younger age groups. The electronic database from a large and comprehensive secondarymental healthcare provider in London was utilized to assess the impact of SMI diagnoses on life expectancy at birth.

Method: People who were diagnosed with SMI (schizophrenia, schizoaffective disorder, bipolar disorder), substance usedisorder, and depressive episode/disorder before the end of 2009 and under active review by the South London andMaudsley NHS Foundation Trust (SLAM) in southeast London during 2007–09 comprised the sample, retrieved by the SLAMCase Register Interactive Search (CRIS) system. We estimated life expectancy at birth for people with SMI and eachdiagnosis, from national mortality returns between 2007–09, using a life table method.

Results: A total of 31,719 eligible people, aged 15 years or older, with SMI were analyzed. Among them, 1,370 died during2007–09. Compared to national figures, all disorders were associated with substantially lower life expectancy: 8.0 to 14.6 lifeyears lost for men and 9.8 to 17.5 life years lost for women. Highest reductions were found for men with schizophrenia (14.6years lost) and women with schizoaffective disorders (17.5 years lost).

Conclusion: The impact of serious mental illness on life expectancy is marked and generally higher than similarly calculatedimpacts of well-recognised adverse exposures such as smoking, diabetes and obesity. Strategies to identify and preventcauses of premature death are urgently required.

Citation: Chang C-K, Hayes RD, Perera G, Broadbent MTM, Fernandes AC, et al. (2011) Life Expectancy at Birth for People with Serious Mental Illness and OtherMajor Disorders from a Secondary Mental Health Care Case Register in London. PLoS ONE 6(5): e19590. doi:10.1371/journal.pone.0019590

Editor: James G. Scott, The University of Queensland, Australia

Received January 24, 2011; Accepted April 1, 2011; Published May 18, 2011

Copyright: ! 2011 Chang et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permitsunrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Funding: The development of the SLAM BRC Case Register was funded by two Capital Awards from the UK National Institute for Health Research and is furthersupported through the BRC Nucleus funded jointly by the Guy’s and St Thomas’ Trustees and South London and Maudsley Special Trustees. C-KC, RH, AF, MB, MHand RS are funded by the National Institute for Health Research (NIHR) Specialist Biomedical Research Centre for Mental Health at the South London andMaudsley NHS Foundation Trust and Institute of Psychiatry, King’s College London. WL is funded by the UK Medical Research Council. The funders had no role instudy design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing Interests: The authors have declared that no competing interests exist.

* E-mail: chin-kuo.chang@kcl.ac.uk

Introduction

Concerns about premature mortality among people with mentaldisorders have been increasing [1–6]. Higher general and specificcauses of mortality in all or specific age groups have been identifiedfor people with serious mental illnesses (SMI, which might includeschizophrenia, schizoaffective disorder, bipolar disorder, anddepressive psychosis) in a range of settings [1,2,7–11], althoughmost report relative risks or rates of mortality rather than lifeexpectancy – which is arguably a more important index for shapingpolicy since it takes premature mortality into account and has beenwidely used for other risk exposures such as smoking, diabetes,obesity, ethnicity and socioeconomic status [12–17]. Of the fewstudies to measure this outcome, clients with SMI from publicmental health agencies in eight US states from 1997 to 2000 werefound to have lower life expectancies of 13 to more than 30 years

compared to the general population (depending on the year andstate) [2], and a loss of 8.8 life years (14.1 years for men and 5.7years for women) was estimated in a comparison between peopletreated for SMI and the general population in Massachusetts, US[1]. A Swedish study using a nationwide hospital discharge registryreported substantial gaps in life expectancy at age 30 for mainmental disorder categories compared to the general population,particularly for functional psychosis other than schizophrenia /affective psychosis (15.9 years lost), substance abuse (15.6 years lost),and organic psychosis (14.8 years lost) among men and organicpsychosis (22.6 years lost), mental retardation (14.7 years lost), andsubstance abuse (18.8 years lost) among women [18]. Preliminarystrategies for preventing premature deaths among people with SMIhave been suggested, emphasizing the management of suicide riskand physical illness, minimum polypharmacy, and improvement ofaccessibility to physical health care [19].

PLoS ONE | www.plosone.org 1 May 2011 | Volume 6 | Issue 5 | e19590

Onset  of  bipolar  disorder  

higher for BP-I (63.3%) and BP-II (73.2%) than for sub-threshold BPD (59.5%) (results available on request). Thesame pattern is found for retrospectively reported meannumber of years in episode (10.3 for BP-I, 11.6 for BP-II, and 6.8 for subthreshold BPD) and mean number oflifetime episodes (77.6 for BP-I, 63.6 for BP-II, and 31.8for subthreshold BPD). More detailed analysis (resultsavailable on request) showed that low persistence of sub-threshold BPD is limited to participants without a his-tory of MDE. The ratio of lifetime manic/hypomanic epi-sodes to total episodes is in the range of 0.5 to 0.6 forrespondents with lifetime BP-I or BP-II and consider-ably higher (0.8) for subthreshold BPD due to the inclu-sion of hypomania in the absence of MDE. Twelve-month ratios are similar to lifetime ratios.

SOCIODEMOGRAPHIC CORRELATES

The sociodemographic correlates of BPD in the NCS-Rare modest in magnitude but fairly consistent across theBPD spectrum. Bipolar disorder is inversely related to ageand educational level and is elevated in previously mar-ried individuals compared with the currently married(only for subthreshold BPD) and in the unemployed-disabled compared with the employed. Bipolar disorderis unrelated to sex, race/ethnicity, and family income (re-sults not presented but available on request).

COMORBIDITY WITHOTHER DSM-IV DISORDERS

Lifetime comorbidity with other DSM-IV/CIDI disor-ders was reported by most respondents with a history ofthreshold (95.8%-97.7%) and subthreshold (88.4%)BPD (Table 2). Odds ratios of BPD with individual dis-orders are uniformly significant and generally higherfor BP-I (5.2-13.7) and BP-II (2.6-16.7) than for sub-threshold BPD (2.2-5.0). An exception is odds ratioswith substance use disorders being much higher forBP-I than for either BP-II or subthreshold BPD. Comor-bidity with 3 or more other disorders is dramatically

higher than comorbidity with only 1 disorder across theBPD spectrum.

SEVERITY OF 12-MONTH CASES

The YMRS 12-month episode severity was rated as se-vere in a higher proportion of BP-I (70.2%) than BP-II(55.4%) or subthreshold BPD (31.5%) (Table 3). Mostmanic/hypomanic episodes were classified as either se-vere or moderate across the BPD spectrum (87.3%-94.6%). The QIDS-SR 12-month severity was rated asmore severe for BP-II (84.0%) than for BP-I (70.5%) orsubthreshold BPD (46.4%). As with mania/hypomania,most MDEs were classified as either severe or moderateacross the spectrum (92.3%-100.0%).

Severe role impairment due to 12-month mania/hypomania was reported by 73.1% of those with 12-month BP-I, 64.6% with BP-II, and 45.9% with subthresh-old BPD (Table 4). Severe role impairment due to 12-month MDE in people with BPD was reported by evenhigher proportions of 12-month cases. As with clinicalseverity, severe role impairment due to MDE was morecommon in participants with BP-II (91.4%) vs BP-I(89.3%) or subthreshold BPD (78.8%). Reports of mod-erate or severe impairment were also more common forMDE (98.6%-100.0%) than for mania/hypomania (87.3%-100%). Impairment was common in all the domains as-sessed using the Sheehan Disability Scale.

TREATMENT

Lifetime treatment of emotional problems by the time ofthe interview was reported by 80.1% of respondents withlifetime BPD (Table 5). A history of treatment is morecommon for BP-I and BP-II (89.2%-95.0%) than for sub-threshold BPD (69.3%), although even the latter is highcompared with other DSM-IV/CIDI disorders.32 Psychia-trists were the most common providers for BP-I (64.9%)and BP-II (62.2%) and general medical professionals forsubthreshold BPD (37.5%).

Summingtreatmentproportionsacrosssectorsshowsmul-tiplesectors tobethenorm,witha2.2-sectoraverageamongpatients (2.7 for BP-I, 2.4 for BP-II, and 1.9 for subthresh-old BPD). Treatment of 12-month BPD was more frequentthan treatment of other DSM-IV/CIDI disorders (67.3% forBP-I, 65.8% for BP-II, and 36.7% for subthreshold BPD).33

Unlike lifetime treatment, nonpsychiatrist mental healthprofessionals were the most common providers (35.4% forBP-I, 33.8% for BP-II, and 20.6% for subthreshold BPD).

Appropriate medication use could be analyzed onlyfor 12-month treatment (Table 6). A significantly higherproportion of patients receiving psychiatric (45.0%) vsgeneral medical (9.0%) treatment received appropriatemedication. Conversely, the proportion of patients whoreceived inappropriate treatment during the past 12months was significantly greater in those who receivedtreatment from general medical physicians (73.1%) vs psy-chiatric specialists (43.4%). A significant gradient wasfound in the proportion of all 12-month cases (ignoringwhether they received treatment) that received appro-priate medication: 25.0% for BP-I, 15.4% for BP-II, and8.1% for subthreshold BPD. The proportion receiving in-

100

80

90

40

10

20

30

70

60

50

0 10 50403020 60 70 80Age, y

Cum

ulat

ive

Life

time

Risk

of D

isor

der,

%

BP-IBP-IISubthreshold BPD

Figure. Cumulative age-at-onset distributions of the DSM-IV/CompositeInternational Diagnostic Interview bipolar disorders (BPDs) in respondentsprojected to develop these disorders in their lifetime. Onset is defined as theage at the first occurrence of either a manic/hypomanic or a majordepressive episode.

(REPRINTED) ARCH GEN PSYCHIATRY/ VOL 64, MAY 2007 WWW.ARCHGENPSYCHIATRY.COM546

©2007 American Medical Association. All rights reserved. at Kings College London, on October 26, 2011 www.archgenpsychiatry.comDownloaded from

USA  (NCS-­‐R)  Arch  Gen  Psychiatry  2007:64:543-­‐552  

0 20 40 60 80

0.0

0.2

0.4

0.6

0.8

1.0

Age of onset (Cardiff cohort)

Age

Pro

babi

lity

of o

nset

18 35

Cardiff  Data  from  Hamshere  et  al  Journal  of  Affec5ve  Disorders  2009:116:23-­‐29  

First  symptoms  

First  episode  

First  treatment  sought  

Bipolar  diagnosis  

Age  (yr):                            17.5                      21.2                  24      30  

Berk  et  al.  (2007)  Journal  of  Affec5ve  Disorders  103:181-­‐186  

Par,cipants  in  Bipolar  Comprehensive  Outcomes  Study  (n=240)  

Delays  in  current  care  

Recurrence  rates  and  number  of  episodes  

Adapted  from  Kessing  et  al.  Bri5sh  Journal  of  Psychiatry  1998:172:23-­‐28  

Denmark:  all  admissions  1971-­‐1993  n~3000  

Cogni,ve  func,on  

Evolution of neuropsychological dysfunction duringthe course of schizophrenia and bipolar disorder

K. E. Lewandowski*, B. M. Cohen and D. Ongur

McLean Hospital and Harvard Medical School, Boston, MA, USA

Background. Neurocognitive dysfunction in schizophrenia (SZ), bipolar (BD) and related disorders represents a core

feature of these illnesses, possibly a marker of underlying pathophysiology. Substantial overlap in domains of

neuropsychological deficits has been reported among these disorders after illness onset. However, it is unclear

whether deficits follow the same longitudinal pre- and post-morbid course across diagnoses. We examine evidence

for neurocognitive dysfunction as a core feature of all idiopathic psychotic illnesses, and trace its evolution from pre-

morbid and prodromal states through the emergence of overt psychosis and into chronic illness in patients with SZ,

BD and related disorders.

Method. Articles reporting on neuropsychological functioning in patients with SZ, BD and related disorders before

and after illness onset were reviewed. Given the vast literature on these topics and the present focus on cross-

diagnostic comparisons, priority was given to primary data papers that assessed cross-diagnostic samples and recent

meta-analyses.

Results. Patients with SZ exhibit dysfunction preceding the onset of illness, which becomes more pronounced in the

prodrome and early years following diagnosis, then settles into a stable pattern. Patients with BD generally exhibit

typical cognitive development pre-morbidly, but demonstrate deficits by first episode that are amplified with

worsening symptoms and exacerbations.

Conclusions. Neuropsychological deficits represent a core feature of SZ and BD; however, their onset and

progression di!er between diagnostic groups. A lifetime perspective on the evolution of neurocognitive deficits in SZ

and BD reveals distinct patterns, and may provide a useful guide to the examination of the pathophysiological

processes underpinning these functions across disorders.

Received 11 August 2009 ; Revised 5 April 2010 ; Accepted 10 April 2010 ; First published online 19 May 2010

Key words : Bipolar, cognitive, longitudinal, premorbid, schizophrenia.

Introduction

Substantial evidence exists of shared genetic liabilitybetween schizophrenia (SZ) and bipolar disorder (BD)(e.g. Tsuang et al. 1980 ; Kendler et al. 1998; Torrey,1999 ; Valles et al. 2000 ; Craddock et al. 2006). Examin-ation of the nature of the development, expression andlongitudinal course of major features of illness mayhelp to illuminate the extent to which specific patho-logical processes are shared or distinct in these dis-orders. Patients with SZ, BD and related disordersexhibit persistent neuropsychological deficits after,and often before, onset of overt illness. Such deficitsmay reflect pathophysiology and neurodevelopment.However, despite substantial overlap in biologicaland clinical features of psychotic disorders, the

longitudinal trajectory of neuropsychological dysfunc-tion across diagnoses remains unclear. Clarification ofthe onset and evolution of neurocognitive deficits inpsychosis is crucial for identifying relevant cognitivephenotypes and clarifying the pathophysiological pro-cesses that are reflected by neuropsychological as-sessment.

Method

The literature examining neuropsychological deficitsin SZ, BD and related disorders is extensive. Thepresent work focused on meta-analyses and cross-diagnostic studies examining neuropsychologicalfunctioning in patients with SZ, BD, schizo-a!ectivedisorder (SZA) and related disorders such as majordepressive disorder with psychosis (MDD) andpsychosis not otherwise specified (NOS). PubMedsearches were performed (most recently on 31 March2010) using the search phrases ‘schizophrenia AND

*Address for correspondence : Dr K. E. Lewandowski, AB347,

McLean Hospital, 115 Mill Street, Belmont, MA 02478, USA.

(Email : klewandowski@mclean.harvard.edu)

Psychological Medicine (2011), 41, 225–241. f Cambridge University Press 2010doi:10.1017/S0033291710001042

REVIEW ARTICLEDiscussion and future directions

Neuropsychological dysfunction represents a measur-able component and characteristic feature of psychoticdisorders. However, the evolution of neurocognitivedysfunction seems to follow distinct courses in SZ andBD (see Fig. 1). Consistent with a neurodevelopmentalmodel (e.g. Weinberger, 1987 ; Andreasen, 1999 ;Marenco & Weinberger, 2000), children who go on todevelop SZ exhibit cognitive deficits relative to theirpeers pre-morbidly. Beginning in the early prodromalphase, patients are found to experience deteriorationin cognitive functioning relative to both their peersand their own prior levels of functioning (Fulleret al. 2002 ; Bilder et al. 2006). Thus, neurocognitivefunctioning in pre-morbid SZ may reflect an abnormalneurodevelopmental process that is present as earlyas objective testing is available, and that lays thefoundation for neurodegeneration or continued mal-development around the time of the SZ prodrome, asreflected by a decline in both relative and absoluteneurocognitive performance. By contrast, patientswith BD exhibit relatively intact cognitive functioningthroughout childhood and adolescence. It is not untilthe time of overt symptom onset that neuropsycho-logical functioning is found to deteriorate, with neuro-psychological deficits detectable by the time of initialdiagnosis. After diagnosis, neuropsychological func-tioning worsens, with acute exacerbations and in con-junction with the presence of psychotic symptoms.This longitudinal course suggests that neurodevelop-mental factors play, at most, a minor role in theemergence of neuropsychological dysfunction inBD, whereas psychopathological factors duringthe course of the disorder itself are associated with

neuropsychological deterioration. Additionally, asdiscussed later, the neurotoxic or neuroprotective roleof medication may a!ect cognitive functioning afterdiagnosis in both SZ and BD.

Although post-onset assessments of neuropsycho-logical functioning reveal patterns of deficits that seemto be similar qualitatively and perhaps quantitativelyacross groups, the pathways by which patient groupsarrive at these points di!er. These di!erences in evol-ution of neuropsychological dysfunction may reflectcontrasting pathogenic pathways in the etiology ofneuropsychological deficits and other symptoms. Suchfindings may provide a theoretical framework forstudying the development of neuropsychologicalsymptoms, and these disorders more generally. Byjuxtaposing the longitudinal course of neuropsycho-logical deficit evolution across disorders, we are betterpositioned to examine their etiological and pathogenicprocesses by capitalizing on what is known to di!erand what is known to overlap, and the timing of each.

We now turn to some methodological and concep-tual debates relevant to this literature.

Identifying the emergence of symptoms pre-morbidly

Perhaps the greatest barrier to clarifying patterns ofdevelopment in psychosis over time is a lack of sys-tematic study beginning pre-morbidly and continuinginto illness phases. Most reports of pre-morbid cogni-tive functioning in children who later develop psy-chosis rely on retrospective reports of pre-morbid IQor review of academic records. Although thesemeasures are useful in estimating pre-morbid cogni-tive functioning, they may be confounded by selectionbias or by a host of variables that interfere with

HealthyBDSZ

Time

Childhood Adolescence Firstepisode

Progression and subsequent acute exacerbations

Leve

l of c

ogni

tive

func

tioni

ng

Fig. 1. Cognitive development in subjects with schizophrenia (SZ), bipolar disorder (BD) and healthy comparisons.

Neuropsychological course in psychosis 233

Substance  Use  Disorders  

•  Substance  use  disorders  comorbidity  high  – Especially  bipolar  I  (OR  8.8;  5.9-­‐13.1)  – bipolar  II  (OR  3.9;  2.7-­‐5.7)  –  NCS-­‐R  study  (n=9282).  Arch  Gen  Psychiatry  2007;64:543-­‐552.  

•  Mood  predicts  later  substance  use  disorders  –  Offspring  of  bipolar  parents.  Duffy  et  al.  J  Affect  Dis  2012;142:57-­‐64  

–  Zurich  Cohort  (n=591):  manic  sx  at  baseline.  Arch  Gen  Psychiatry  2008;65:47-­‐52.  

Impact  of  delayed  diagnosis  

529  outpa,ents  with  bipolar  disorder  Delay  to  first  treatment  associated  with:    More  ,me  depressed    Greater  severity  of  episodes    More  episodes    Fewer  days  euthymic  

 Post  et  al.  J  Clin  Psychiatry  2010:71:864-­‐872  

psychiatric medications prescribed prior to receivingtheir current diagnosis.

3. Results

Of the 240 total participants in the BCOS study, oneparticipant withdrew consent and 218 participantsanswered the questionnaire. This cohort consisted of125 females and 93 males. At entry into the BCOS study163 participants had a current diagnosis of bipolardisorder and 55 participants had a current diagnosis ofschizoaffective disorder. The mean age of participantswas 41.8 years (SD 12.7 years).

Any symptoms of mental illness were first experi-enced at 17.5 years (median; Inter Quartile Range (IQR)13.8–24.3; n=216) and mood swings at 18.0 years(median; IQR 14–25; n=197). Symptoms of depressionwere first experienced at 18.0 years (median; IQR 14–25; n=197), a full episode of depression at 21.2 years(median; IQR 17–28.5; n=200), symptoms of mania at21.0 years (median; IQR 16.8–29.5; n=212) and a fullepisode of mania at 24.1 years (median; IQR 19–30.5;n=205). Medical treatment was first sought at24.0 years (median; IQR 19–31.5; n=217). Participantsfirst received a diagnosis of Bipolar Disorder orSchizoaffective Disorder at 30 years (median; IQR23–37.3; n=215) (Fig. 1). There was a trend associatingearlier onset of illness with longer delays in diagnosis ofbipolar or schizoaffective disorder (Fig. 2). The meandelay from first experiencing any symptoms of mentalillness to diagnosis was 16.3±10.5 years (mean±SD)for participants aged 3 to 15.9 years, 9.9±9.5 years(mean±SD) for participants aged 16 to 25.9 years, 7.3±9.9 years (mean±SD) for participants aged 26 to35.9 years and 2.7±5.1 years (mean±SD) for partici-pants aged >35.9 years.

Having had a diagnosis prior to a final diagnosis ofBipolar Disorder or Schizoaffective Disorder wasreported by 120 of 218 participants who answered thisquestion. The most commonly reported prior diagnosiswas depression (26.6% of n=218). A prior diagnosis of

depression was retained by 58 participants for 7.6±8.7 years (mean±SD). A prior diagnosis of schizo-phrenia was retained by 27 participants for 8.6±6.5 years (mean±SD). A prior diagnosis of bipolardisorder (final diagnosis schizoaffective disorder) wasretained by 13 participants for 5.2±4.6 years (mean±SD). A prior diagnosis of postnatal depression orpostnatal psychosis was retained by 12 participants for3.2±4.1 years (mean±SD). A prior diagnosis ofschizoaffective disorder (final diagnosis bipolar dis-order) was retained by 7 participants for 4.5±3.0 years(mean±SD). A prior diagnosis of borderline personalitydisorder was retained by 6 participants for 5.8±5.6 years (mean±SD).

Having been treated with psychiatric medicationsprior to a final diagnosis of Bipolar Disorder orSchizoaffective Disorder was reported by 121 of the218 participants who answered this question. Partici-pants were also questioned whether they had beenadherent to these medications and whether in theiropinions the medications were useful. Participantsreported 50 antidepressant exposures to SSRIs of 1.8±2.6 years (mean±SD), 25 exposures to TCAs of 2.1±2.8 years (mean±SD), 9 exposures to MAOIs of 1.2±0.9 years (mean ± SD), 13 exposures to other

Fig. 1. Timeline for the median age (years) of illness milestones for 207 study participants with bipolar disorder or schizoaffective disorder.

Fig. 2. Graph of age of onset of the first symptoms of mental illnessversus years of delay prior to a confirmed diagnosis of bipolar orschizoaffective disorder for participants in the BCOS study (n=213).Longer delays in diagnosis for an earlier age of onset are demonstrated.

183M. Berk et al. / Journal of Affective Disorders 103 (2007) 181–186

Berk  et  al.  (2007)  Journal  of  Affec5ve  Disorders  103:181-­‐186  

Australia:  Par,cipants  in  Bipolar  Comprehensive  Outcomes  Study  (n=240)  

J Clin Psychiatry 71:7, July 2010 868

Effects of Early Bipolar Onset and Treatment Delay

and delay to treatment variables separately. The length of delay to first treatment was correlated with each measure of outcome in the first year of follow-up in an adverse direction (Table 2; left-side columns). The strongest relationships were with the total number of episodes, both measures of depres-sion, days euthymic, and days of ultradian cycling. Only the correlations with severity of mania and days manic would lose significance with a Bonferroni correction.

Correlations with age at onset revealed relationships with a similar magnitude. However, correlations with severity of depression and days depressed were not significant, and the correlation with days manic was not significant after Bonferroni correction. Again, the correlation with number of episodes was the strongest.

Next, multiple linear regressions were run to examine the independent contributions of age at onset and delay to first treatment to outcome in the first year, as this year had the largest N and would be less affected by the type of prospective treatment received in the network. Table 2 (right-side col-umns) shows the standardized regression coefficients from each regression with a main-effects model. Interactions were not included since that reduced tolerance to unacceptable levels without explaining additional variance. The delay to first treatment was independently related to the duration and severity of depression, days well, days of ultradian cycling, and total number of episodes. Early age at onset was an in-dependent contributor only to the total number of episodes observed prospectively.

DISCUSSION

These data extend previous reports indicating that early age at onset is a poor prognosis factor for outcome retrospectively reported by adult outpatients with bipolar disorder studied in research networks.2,3 Our previous short report confirmed these self-report findings with prospective

clinician ratings for 1 year,2 and the current study extends many of these findings over 4 years of prospective follow-up. In addition, the new data provided here supplement previ-ous observations in several different ways. As illustrated in Figure 2, the striking inverse relationship between earlier age at onset of illness and longer delay to first treatment oc-curred whether the first episode was manic, depressive, or both in the same year. As illustrated in Table 2, we found that the duration of the delay to first treatment had an inde-pendent relationship to the prospective measures of severity of depression, number of days depressed, number of days euthymic, number of episodes, and days of ultradian cycling. Age at onset had an independent contribution only to the number of episodes observed prospectively. Whether early intervention to shorten the delays to first treatment could alter this adverse course of illness in adulthood, or wheth-er early onset is a harbinger of worse course regardless of intervention, remains to be studied.

Our initial report, examining the first year, found that an early age at onset was a poor prognostic factor for all 7 variables examined: severity and duration of mania and of depression, days euthymic, number of episodes, and days of ultradian cycling. Extending the analysis to cover 4 years found this relationship to still exist for severity and dura-tion of depression and days euthymic. The differences seen in the first year for severity of mania and days euthymic as a function of the childhood-onset group were attenuated and no longer present with additional time and treatment in the network. When we examined all 4 years, there were no longer any significant group differences in days manic or days of ultradian cycling. This finding suggests that the early age-at-onset grouping is associated with a poor prognosis for depressive symptoms in adulthood that even dedicated prospective treatment over 4 years does not fully alleviate.

Figure 2. Relationship Between Age at Onset of Bipolar Illness and Years of Delay to First Treatment According to Polarity of First Episodea

aDelay to first treatment for mania or depression remains inversely correlated to age at onset of illness regardless of the polarity of the first episode, ie, depression with dysfunction, mania (or hypomania), or both episodes occurring in the same year.

20

15

10

5

0

Del

ay to

Firs

t Tre

atm

ent,

y

Age at Onset

DepressionManiaBoth depression and mania in same year

< 13 y 13–18 y 19–29 y ! 30 y

Table 2. Treatment Delay Is an Independent Correlate of Poor Prospective Outcome in Adults With Bipolar Disordera

Adult Outcome Measure

Pearson Correlations Linear Regressions

Age at OnsetTreatment

DelayAge at Onset

Treatment Delay

r Pb r Pb ! P ! Pb

DepressionSeverity 0.08 .07 "0.15 .002 0.02 .69 "0.14 .01Days "0.08 .08 0.13 .006 "0.03 .64 0.12 .02

ManiaSeverity "0.13 .004 0.11 .02 "0.10 .06 0.07 .22Days "0.10 .03 0.10 .04 "0.07 .21 0.07 .22

Euthymic days

0.17 < .001 "0.21 < .001 0.09 .11 "0.17 < .001

Ultradian cycling days

"0.16 < .001 0.17 < .001 "0.10 .06 0.12 .02

No. of episodes (DSM-IV)

"0.22 < .001 0.28 < .001 "0.12 .02 0.22 < .001

aBoth earlier age at onset of bipolar disorder and longer time delay to first treatment for mania or depression correlate significantly with multiple prospective measures in the first-year follow-up of a more adverse outcome in adulthood (left columns). Independently of age at onset, treatment delay is a significant correlate of a poor outcome on all measures except manic severity and duration.

bBoldface type indicates statistical significance.

Post  et  al.  (2010)  Journal  of  Clinical  Psychiatry  71:864-­‐872  

Interna,onal:  Par,cipants  in  Stanley  Founda,on  Bipolar  Network  (n=529)  

What  should  we  all  be  doing?  

•  Be5er  diagnosis  –  earlier  recogni,on  

•  Be5er  treatment  –  esp.  preventa,ve    

First  symptoms  

First  episode  

First  treatment  sought  

Bipolar  diagnosis  

Age  (yr):                            17.5                        21.2                    24            30  

Berk  et  al.  (2007)  Journal  of  Affec5ve  Disorders  103:181-­‐186  

Par,cipants  in  Bipolar  Comprehensive  Outcomes  Study  (n=240)  

Delays  in  current  care  

In  contact  but  not  yet  bipolar…  

SLaM  cohort  

•  pseudonymised  electronic  health  records    •  Cohort  of  1,440  pa,ents:  –  (i)  Age:  16  –  65  years  –  (ii)  First  referral  to  SLaM  between  01/01/2007  and  31/05/2012  

–  (iii)  Subsequent  diagnosis  of  bipolar  disorder  

Is  there  a  role  for  self-­‐report  screening  instruments?  

•  MDQ  (Mood  Disorder  Ques,onnaire)  –  Includes  13-­‐item  yes/no.  

•  HCL-­‐32  (Hypomania  Checklist)  –  Includes  32-­‐item  yes/no.  – Shorter  versions  in  development  

•  BSDS  (Bipolar  Spectrum  Diagnos,c  Scale)  – Ra,ng  how  well  a  descrip,ve  paragraph  fits  

Hirschfeld,  R.  M.,  et  al.  American  Journal  of  Psychiatry,  2000;157,  1873–1875.  Angst,  J.,  et  al.  Journal  of  Affec5ve  Disorders,  2005;88,  217–233.    Ghaemi,  S.  N.  et  al.  Journal  of  Affec5ve  Disorders,  2005;84,  273–277  

Recogni,on  of  bipolar  disorder  

Bipolar disorder is a complex mood disorder that can be viewed as aspectrum condition.1,2 A growing body of research suggests that it isunderrecognised in clinical practice and often misdiagnosed asrecurrent major depressive disorder.3–8 Estimates of the meandelay between first onset of mood symptoms and receiving acorrect bipolar diagnosis are in the region of 10 years.9–11

Certain subgroups of individuals with depression, such asthose with early-onset depression12,13 and those with severe ortreatment-refractory depression,14 appear to have the highest ratesof unrecognised bipolar disorder. However, many of the studiesthat have identified underdiagnosis of bipolar disorder havebeen carried out in secondary or tertiary care settings and there haveas yet been no studies in the UK assessing the likely prevalence ofunrecognised bipolar disorder in primary care patients who have aworking diagnosis of unipolar depression.

This study had three main aims. First, to estimate theproportion of primary care patients with a working diagnosis ofunipolar depression who satisfy DSM–IV15 diagnostic criteriafor bipolar disorder (either bipolar I disorder, bipolar II disorderor bipolar disorder not otherwise specified). Second, to test thepotential usefulness of two screening instruments for bipolardisorder – the Hypomania Checklist (HCL–32)16 and BipolarSpectrum Diagnostic Scale (BSDS)17 – within a primary caresample. Third, to assess whether those participants with majordepressive disorder who screen false positive for bipolar disorderon the HCL–32 or BSDS questionnaires (that is, participants withmajor depressive disorder and with a history of subdiagnosticmanic symptoms) differ from participants with major depressivedisorder but with no or only minimal history of manic symptoms

in terms of their clinical course, psychosocial functioning andquality of life.

Method

Recruitment

This study was approved by the South East Wales Research EthicsCommittee, Cardiff and Vale University Health Board, and Cardiff,Merthyr Tydfil and Newport Local Health Boards (primary carelocal health boards). Our aim was to collect detailed diagnostic,clinical, psychosocial functioning and quality of life data on arepresentative sample of individuals from primary care who hada current working diagnosis of unipolar depression. In order tosample people from a range of socioeconomic backgrounds, weinvited 45 general practices in three local health boards in SouthWales to take part in this study (27 practices from Cardiff LocalHealth Board, 9 from Merthyr Tydfil Local Health Board and9 from Newport Local Health Board). Eleven practices agreed totake part (seven from Cardiff, three from Merthyr and one fromNewport). From their patient databases, practice managersgenerated lists of individuals who had at least one of eight possiblecodes for depression recorded by their general practitioner (GP)within the past 5 years. These codes were: ‘recurrent depressivedisorder’; ‘depressive episode’; ‘depressive disorder NEC (notelsewhere specified)’; ‘recurrent major depressive episode’; ‘singlemajor depressive episode’; ‘neurotic (reactive) depression’;‘depressed’; and ‘depression NOS (not otherwise specified)’.

When large numbers of potentially eligible participants wereidentified within a single practice, a maximum of 400 wereselected using computer-generated random number lists. In total,3117 patients were identified as potentially eligible from the

49

Unrecognised bipolar disorder in primary carepatients with depression{

Daniel J. Smith, Emily Griffiths, Mark Kelly, Kerry Hood, Nick Craddock and Sharon A. Simpson

BackgroundBipolar disorder is complex and can be difficult to diagnose.It is often misdiagnosed as recurrent major depressivedisorder.

AimsWe had three main aims. To estimate the proportion ofprimary care patients with a working diagnosis of unipolardepression who satisfy DSM–IV criteria for bipolar disorder.To test two screening instruments for bipolar disorder (theHypomania Checklist (HCL–32) and Bipolar SpectrumDiagnostic Scale (BSDS)) within a primary care sample. Toassess whether individuals with major depressive disorderwith subthreshold manic symptoms differ from thoseindividuals with major depressive disorder but with no orlittle history of manic symptoms in terms of clinical course,psychosocial functioning and quality of life.

MethodTwo-phase screening study in primary care.

ResultsThree estimates of the prevalence of undiagnosed bipolardisorder were obtained: 21.6%, 9.6% and 3.3%. The HCL–32

and BSDS questionnaires had quite low positive predictivevalues (50.0 and 30.1% respectively). Participants with majordepressive disorder and with a history of subthreshold manicsymptoms differed from those participants with no or littlehistory of manic symptoms on several clinical features andon measures of both psychosocial functioning and quality oflife.

ConclusionsBetween 3.3 and 21.6% of primary care patients withunipolar depression may have an undiagnosed bipolardisorder. The HCL–32 and BSDS screening questionnairesmay be more useful for detecting broader definitions ofbipolar disorder than DSM–IV-defined bipolar disorder.Subdiagnostic features of bipolar disorder are relativelycommon in primary care patients with unipolar depressionand are associated with a more morbid course of illness.Future classifications of recurrent depression should includedimensional measures of bipolar symptoms.

Declaration of interestD.J.S. has received honoraria for speaking at educationalmeetings organised by AstraZeneca and Lilly.

The British Journal of Psychiatry (2011)199, 49–56. doi: 10.1192/bjp.bp.110.083840

{See editorial, pp. 3–4, this issue.

Bipolar disorder is a complex mood disorder that can be viewed as aspectrum condition.1,2 A growing body of research suggests that it isunderrecognised in clinical practice and often misdiagnosed asrecurrent major depressive disorder.3–8 Estimates of the meandelay between first onset of mood symptoms and receiving acorrect bipolar diagnosis are in the region of 10 years.9–11

Certain subgroups of individuals with depression, such asthose with early-onset depression12,13 and those with severe ortreatment-refractory depression,14 appear to have the highest ratesof unrecognised bipolar disorder. However, many of the studiesthat have identified underdiagnosis of bipolar disorder havebeen carried out in secondary or tertiary care settings and there haveas yet been no studies in the UK assessing the likely prevalence ofunrecognised bipolar disorder in primary care patients who have aworking diagnosis of unipolar depression.

This study had three main aims. First, to estimate theproportion of primary care patients with a working diagnosis ofunipolar depression who satisfy DSM–IV15 diagnostic criteriafor bipolar disorder (either bipolar I disorder, bipolar II disorderor bipolar disorder not otherwise specified). Second, to test thepotential usefulness of two screening instruments for bipolardisorder – the Hypomania Checklist (HCL–32)16 and BipolarSpectrum Diagnostic Scale (BSDS)17 – within a primary caresample. Third, to assess whether those participants with majordepressive disorder who screen false positive for bipolar disorderon the HCL–32 or BSDS questionnaires (that is, participants withmajor depressive disorder and with a history of subdiagnosticmanic symptoms) differ from participants with major depressivedisorder but with no or only minimal history of manic symptoms

in terms of their clinical course, psychosocial functioning andquality of life.

Method

Recruitment

This study was approved by the South East Wales Research EthicsCommittee, Cardiff and Vale University Health Board, and Cardiff,Merthyr Tydfil and Newport Local Health Boards (primary carelocal health boards). Our aim was to collect detailed diagnostic,clinical, psychosocial functioning and quality of life data on arepresentative sample of individuals from primary care who hada current working diagnosis of unipolar depression. In order tosample people from a range of socioeconomic backgrounds, weinvited 45 general practices in three local health boards in SouthWales to take part in this study (27 practices from Cardiff LocalHealth Board, 9 from Merthyr Tydfil Local Health Board and9 from Newport Local Health Board). Eleven practices agreed totake part (seven from Cardiff, three from Merthyr and one fromNewport). From their patient databases, practice managersgenerated lists of individuals who had at least one of eight possiblecodes for depression recorded by their general practitioner (GP)within the past 5 years. These codes were: ‘recurrent depressivedisorder’; ‘depressive episode’; ‘depressive disorder NEC (notelsewhere specified)’; ‘recurrent major depressive episode’; ‘singlemajor depressive episode’; ‘neurotic (reactive) depression’;‘depressed’; and ‘depression NOS (not otherwise specified)’.

When large numbers of potentially eligible participants wereidentified within a single practice, a maximum of 400 wereselected using computer-generated random number lists. In total,3117 patients were identified as potentially eligible from the

49

Unrecognised bipolar disorder in primary carepatients with depression{

Daniel J. Smith, Emily Griffiths, Mark Kelly, Kerry Hood, Nick Craddock and Sharon A. Simpson

BackgroundBipolar disorder is complex and can be difficult to diagnose.It is often misdiagnosed as recurrent major depressivedisorder.

AimsWe had three main aims. To estimate the proportion ofprimary care patients with a working diagnosis of unipolardepression who satisfy DSM–IV criteria for bipolar disorder.To test two screening instruments for bipolar disorder (theHypomania Checklist (HCL–32) and Bipolar SpectrumDiagnostic Scale (BSDS)) within a primary care sample. Toassess whether individuals with major depressive disorderwith subthreshold manic symptoms differ from thoseindividuals with major depressive disorder but with no orlittle history of manic symptoms in terms of clinical course,psychosocial functioning and quality of life.

MethodTwo-phase screening study in primary care.

ResultsThree estimates of the prevalence of undiagnosed bipolardisorder were obtained: 21.6%, 9.6% and 3.3%. The HCL–32

and BSDS questionnaires had quite low positive predictivevalues (50.0 and 30.1% respectively). Participants with majordepressive disorder and with a history of subthreshold manicsymptoms differed from those participants with no or littlehistory of manic symptoms on several clinical features andon measures of both psychosocial functioning and quality oflife.

ConclusionsBetween 3.3 and 21.6% of primary care patients withunipolar depression may have an undiagnosed bipolardisorder. The HCL–32 and BSDS screening questionnairesmay be more useful for detecting broader definitions ofbipolar disorder than DSM–IV-defined bipolar disorder.Subdiagnostic features of bipolar disorder are relativelycommon in primary care patients with unipolar depressionand are associated with a more morbid course of illness.Future classifications of recurrent depression should includedimensional measures of bipolar symptoms.

Declaration of interestD.J.S. has received honoraria for speaking at educationalmeetings organised by AstraZeneca and Lilly.

The British Journal of Psychiatry (2011)199, 49–56. doi: 10.1192/bjp.bp.110.083840

{See editorial, pp. 3–4, this issue.

Primary  care  pa,ents  with  working  diagnosis  of  unipolar  depression  Screening  instruments  (HCL-­‐32  &  BSDS)    

     High  score        Low  score  

11 practices and all were invited to take part in this study. Theinvitation sheet about the study was sent to individuals bypractice managers on behalf of the research team and includedthe HCL–32 and BSDS questionnaires. Invited participants wereasked to return the completed questionnaires if they wereinterested in taking part in the study. A reminder invitation packwas sent by practice managers to those who had not replied after8 weeks.

The recruitment process is outlined in Fig. 1. In total, 576individuals (18.5% of those invited) volunteered to take partand returned completed HCL–32 and BSDS questionnaires. Inorder to compare volunteers and non-volunteers, each practicemanager used computer-generated random number lists toselect a small sample of 55 volunteers and 55 non-volunteers(5 volunteers and 5 non-volunteers from each practice) andperformed a case-file review of age, gender, time since firstdiagnosis of depression, whether currently taking antidepressantmedication, whether taking other medication, and whetherdiagnosed with a comorbid medical disorder (Table 1).

The 576 participants who returned HCL–32 and BSDSquestionnaires were divided into two groups: a group of ‘highscorers’ (defined as scoring either 14 or more on the HCL–32 or13 or more on the BSDS; n= 411) and a group of ‘low scorers’(defined as scoring less than 14 on the HCL–32 and less than13 on the BSDS; n= 165) (Fig. 1). This approach is outlined inthe recruitment flow chart (Fig. 1). The choice of these thresholdswas informed by several previous studies that suggested that 14 ormore on the HCL–32 and 13 or more on the BSDS representedreasonable cut-offs on each questionnaire for reliably distinguishingbetween bipolar disorder and major depressive disorder.16–20

From the 576 volunteers, all of the ‘low scorers’ group(n= 165) and 50% of the ‘high scorers’ group (selected usingcomputer-generated random number lists; n= 205) were invitedto take part in a comprehensive diagnostic and clinical assessment.In total, 154 (41.6%) of these 370 eligible participants wereassessed at interview. To test the representativeness of thissample, the 154 participants who were interviewed werecompared with the 216 participants who declined to beinterviewed (Table 2).

Assessments

The comprehensive diagnostic and clinical assessment wascarried out by a research psychologist and included: the MiniInternational Neuropsychiatric Interview (MINI) to obtainDSM–IV diagnoses;21 a structured assessment of socio-demographics, medical history, current medication and family

50

Smith et al

3117 eligible individuals invited to take part

576 returned HCL–32 and BSDS questionnaires

Not interviewed(n = 113)

Not interviewed(n = 103)

‘High scorers’(n = 411)

50% randomly invitedto interview

(n = 205)

Interviewed(n = 92)

Majordepressive

disorder(n = 61)

‘Low scorers’(n = 165)

All invitedto interview

(n = 165)

Interviewed(n = 62)

Majordepressive

disorder(n = 55)

Bipolardisorder(n = 26)

Notmajor

depressivedisorder

(n = 4)

Bipolardisorder

(n = 3)

Notmajor

depressivedisorder

(n = 5)

6

6

6

6

66 6

6

6 6 6

6

6

6

7

8

Bipolar I disorder (n = 2)Bipolar II disorder (n = 14)Bipolar disorder – NOS(n = 10)

Bipolar I disorder (n = 0)Bipolar II disorder (n = 1)Bipolar disorder – NOS(n = 2)

Fig. 1 Recruitment flow chart.

‘High scorers’ were defined as having either a score of 14 or more on the HypomaniaChecklist–32 (HCL–32) or 13 or more on the Bipolar Spectrum Diagnostic Scale (BSDS).‘Low scorers’ were defined as having a score of less than 14 on the HCL–32 and lessthan 13 on the BSDS. NOS, not otherwise specified.

Table 1 Characteristics of a random subsample of volunteers v. non-volunteers

Volunteers (n= 55) Non-volunteers (n= 55) P

Age, years: mean (s.d.) 42.8 (13.0) 44.4 (11.1) 0.49a

Gender ratio, male:female 17:38 18:37 0.84b

Time since first diagnosis, months: median (range) 60 (4–492) 48 (3–540) 0.51c

Currently taking antidepressants, n (%) 32 (58.2) 18 (32.7) 0.01b

Currently taking other medications, n (%) 35 (63.6) 27 (49.1) 0.12b

At least 1 comorbid medical disorder, n (%) 28 (50.9) 22 (40.0) 0.29b

a. Independent t-test.b. w2-test.c. Mann–Whitney U-test.

Table 2 Comparison of interviewees v. those who were not interviewed

Interviewed (n= 154) Not interviewed (n= 216) P

Age, years: mean (s.d.) 44.0 (12.5) 40.9 (13.7) 0.03a

Male, n (%) 53 (34.4) 58 (26.9) 0.12b

Ethnicity, White: n (%) 118 (76.6) 155 (71.8) 0.17b

Hypomania Checklist–32 score, mean (s.d.) 12.83 (7.14) 14.09 (7.46) 0.10a

Bipolar Spectrum Diagnostic Scale score, mean (s.d.) 11.77 (6.14) 12.66 (6.45) 0.19a

a. Independent t-test.b. w2-test.

What  should  we  all  be  doing?  

•  Be5er  diagnosis  –  earlier  recogni,on  

•  Be5er  treatment  –  esp.  preventa,ve    

Time  to  appropriate  treatment  •  France  (Drancourt  et  al.  2013)  – Median  four  years  from  first  admission  to  ‘mood  stabiliser’  (n=501)  •  [atypical  an,psycho,c,  lithium,  an,convulsant]  

Figure 2 shows the possible pathway from BDonset to guideline-recognized treatment for thewhole study sample, but we then explored keyclinical milestones for subgroups defined accordingto polarity of first episode – mania, depression orhypomania.For the manic onset subgroup (n = 135), first

BD episode typically occurred at about 24 years ofage, with hospitalization and treatment with psy-chotropics occurring simultaneously about 2 yearslater. First suicide attempt occurred about 3 yearsafter that (median age, 29 years), preceding receiptof first mood stabilizer treatment by about 3 years.In the subgroups with depressive (n = 298) orhypomanic first episodes (n = 50), it is notablethat these occur earlier than mania, and suicideattempts typically occur prior to the introductionof any form of treatment. The longest DUB was inthose presenting with hypomania (median,14.5 years) rather than depression (median DUBof 13 years) or mania (median DUB of 8 years).When diagnostic subtype was taken into account,the median DUB in BD-II cases presenting withdepression was about 14 years compared withabout 11 years for BD-I cases (data available onrequest) (Fig. 3).

Discussion

We report here on the largest clinical study ofDUB undertaken so far. It shows that, in a Frenchsample of over 500 BD-I and BD-II patients, themean DUB was 9.6 (±9.7) years, with a median of6 years. It is noteworthy that this finding is similarto those reported up to 25 years ago in interna-tional studies of the delay in or mis-diagnosis ofBD and ⁄or of delayed treatment of BD (9–12, 15).A di!culty in making direct comparisons betweenstudies is that some focus on time from firstsymptoms to diagnosis, others target time betweenfirst episode and any treatment, while others reportdelay in prescription of a mood stabilizer.However, these di"erent approaches actuallyreflect di"erent aspects of the same problem (e.g.latency to help-seeking, latency to diagnosis andlatency to appropriate treatment). Sadly, it appearsthat whatever definition is used, delayed diagnosisof BD and ⁄or late introduction of guideline-defined appropriate treatments remains a universalunmet need, whatever the country, the mentalhealthcare system and ⁄or its accessibility.Previous, smaller studies report that the DUB is

significantly longer in BD-II patients, particularlywomen (26, 27). Our findings are consistent withthe well-documented high rate of undiagnosedhypomania or failure to consider the clinicalimplications of such a presentation (28). Althoughwomen are more likely to seek help and to accesshealthcare services (29–31), we found that theywait 2 years longer than men before being pre-scribed a mood stabilizer, suggesting no improve-ment in this gender bias over the last decade (26,32).We demonstrated that onset characteristics sig-

nificantly influenced the DUB. An early onset (age£21 years) is associated with a longer DUB, whichrepresents the most replicated result in the litera-ture (9, 10, 12, 15, 31, 33–35). While this may

Fig. 1. Duration of untreated bipolar disorder (DUB)according to gender and bipolar disorder sub-type.

Fig. 2. Key clinical milestones (meanage in years ± SD) during the Dura-tion of Untreated Bipolar Disorder(DUB) for all study participants.

Drancourt et al.

140

The  role  of  ‘Early  Interven,on’  teams  

•  [Already  part  of  EI  services]  NICE  guidance  for  Bipolar  Disorder  (CG38)    para  1.3.2.7  'Early  interven5on  services  for  people  with  psychosis  should  be  available  to  people  with  bipolar  disorder...'  

Tradi,onal  Early  Interven,on  for  Psychosis  

•  Minimising  dura,on  of  untreated  psychosis  •  Iden,fica,on  in  the  prodromal  period  •  Mul,disciplinary  approach  – Maximise  engagement  – Maximise  adherence  

•  ‘zero  tolerance  for  side  effects’  

First  Episode  Community  Teams  And  First  Episode  Inpa,ent  Unit  

High  Risk  for  Schizophrenia  

Schizophrenia  /    Non-­‐affec,ve  

Bipolar  /  Affec,ve  

baseline 18 months

EPPICFirst psychotic illness

0100

200

300

400

bipolar

SczAffective

Schizophrenia

other psychosis

13

4

9

Schimmelmann  et  al.  2005  J  Clin  Psychiatry  66:1239-­‐1246  

community  

baseline 24 months

Harvard-McLean First Episode ProjectFirst psychotic illness

0100

200

300

400

500

bipolar

MDD

other psychosis

SczAffective13

16

8

Salvatore  et  al.  2009  J  Clin  Psychiatry  70:458-­‐466  

inpa,ent  

Bipolar  II  /    Bipolar  Spectrum  

First  Episode  Psychosis  

At  Risk  Mental  State  

Schizophrenia  /    Non-­‐affec,ve  

Bipolar  /  Affec,ve  

Bipolar  Disorder  care  within  EI  services  

•  Minimising  dura,on  of  untreated  bipolar  •  Iden,fica,on  in  the  prodromal  period  (?)  •  Mul,disciplinary  approach  – Maximise  engagement  – Maximise  adherence  

•  ‘zero  tolerance  for  side  effects’  

Future  direc,ons  

•  Prodromal  or  ‘Bipolar  At  Risk’  state  – Predic,on  of  onset/conversion  – Preven,on  of  onset/conversion  

First  symptoms  

First  episode  

First  treatment  sought  

Bipolar  diagnosis  

Age  (yr):                            17.5                      21.2                    24          30  

Berk  et  al.  (2007)  Journal  of  Affec5ve  Disorders  103:181-­‐186  

Par,cipants  in  Bipolar  Comprehensive  Outcomes  Study  (n=240)  

Delays  in  current  care  

First  Episode  Community  Teams  And  First  Episode  Inpa,ent  Unit  

High  Risk  for  Schizophrenia  

High  Risk  for  Bipolar  

Schizophrenia  /    Non-­‐affec,ve  

Bipolar  /  Affec,ve  

Early  Interven,on  in  Bipolar  Disorder  

•  Why  do  we  need  to  intervene  earlier  in  bipolar  disorder?  – Diagnos,c  delays  – How  does  bipolar  disorder  fit  within  specialist  ‘Early  Interven,on’  services?  

•  What  are  the  future  direc,ons?