Evidence profile - non-pharmacological treatment of depression

Forest plot identifierDescriptionStudy IdsQuality of included studiesConsistencyDirectnessOther factorsN treatment group/N controlEffect sizeAbsolute statistics (where appropriate)Likelihood of clinically important effectOverall quality
All therapies versus standard care/waitlist
RMnpta 01.01Depression above threshold at endpointCooper2003 UK
Honey2002 UK
O'Hara2000 US
Prendergast2001 Aus
Holden1989 UK
Wickberg1996B Swe
Chen2000 Taiw
Armstrong2003B Aus
SIGN: 1+Significant heterogeneity: random effects model used (−1)RelevantNone244/242RR (random effects) = 0.57 (0.43, 0.76)Very likely (favouring treatment)Moderate
RMnpta 01.02Mean endpoint scoresCooper2003 UK
Honey2002 UK
Milgrom2005 Aus
Ohara2000 US
Prendergast2001 Aus
Armstrong2003B Aus
SIGN: 1+Significant heterogeneity: random effects model used (−1)RelevantNone174/171SMD (random effects) = − 0.61 (−1.07, − 0.14)Likely (favouring treatment)Moderate
RMnpta 01.03Mean change scoresChen2000 TaiwSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)30/30SMD = −0.71 (−1.23, −0.19)Likely (favouring treatment)Low
Dataset by therapy
Psychological therapies
RMnpta 01.04Depression above thresholdCooper2003 UK
Honey2002 UK
O'Hara2000 US
Prendergast2001 Aus
SIGN: 1+Significant heterogeneity: random effects model used (−1)RelevantNone150/154RR (random effects) = 0.61 (0.4, 0.92)Likely (favouring treatment)Moderate
RMnpta 01.06Mean endpoint scoresCooper2003 UK
Honey2002 UK
Milgrom2005 Aus
O'Hara2000 US
Prendergast2001 Aus
SIGN: 1+Significant heterogeneity: random effects model used (−1)RelevantNone164/161SMD (random effects) = − 0.49 (−0.95, − 0.02)Likely (favouring treatment)Moderate
Listening visits
RMnpta 01.05Depression above thresholdCooper2003 UK
Holden1989 UK
Wickberg1996B Swe
SIGN: 1+No significant heterogeneityRelevantNone98/100RR = 0.61 (0.45, 0.83)Very likely (favouring treatment)High
RMnpta 01.06Mean endpoint scoresCooper2003 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)47/50SMD = −0.26 (−0.66, 0.14)InconclusiveLow
Social support
RMnpta 01.04Depression above thresholdChen2000 TaiwSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)34/30RR = 0.69 (0.42, 1.13)Likely (favouring treatment)Low
RMnpta 01.07Mean change scoresChen2000 TaiwSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)30/30SMD = −0.71 (−1.23, −0.19)Likely (favouring treatment)Low
Group exercise with social support
RMnpta 01.04Depression above thresholdArmstrong2003B AusSIGN: 1+N/ARelevantSparse/imprecise data (−1)10/10RR = 0.07 (0, 1.03)InconclusiveLow
RMnpta 01.06Mean change scoresArmstrong2003B AusSIGN: 1+N/ARelevantSparse/imprecise data (−1)10/10SMD = −1.64 (−2.68, −0.59)Very likely (favouring treatment)Moderate
By baseline diagnosis method
DSM or similar
RMnpta 01.08Depression above thresholdArmstrong2003B Aus
Cooper2003 UK
Holden1989 UK
O'Hara2000 US
Prendergast2001 Aus
SIGN: 1+NoneRelevantStrong effect (+1)163/166RR = 0.46 (0.36, 0.60)Very likely (favouring treatment)High
RMnpta 01.010Mean endpoint scoresArmstrong2003B Aus
Cooper2003 UK
Milgrom2005 Aus
O'Hara2000 US
Prendergast2001 Aus
SIGN: 1+NoneRelevantNone151/149SMD (random effects) = − 0.66 (−1.21, − 0.11)Likely (favouring treatment)Moderate
EPDS >= 12, BDI >= 10 or similar
RMnpta 01.08Depression above thresholdChen2000 Taiw
Honey2002 UK
Wickberg1996B Swe
SIGN: 1+Significant heterogeneity: random effects model used (−1)RelevantNone81/76RR (random effects) = 0.72 (0.49, 1.05)Likely (favouring treatment)Moderate
RMnpta 01.011Mean endpoint scoresHoney2002 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)23/22SMD = −0.36 (−0.95, 0.23)InconclusiveLow
RMnpta 01.010Mean change scoresChen2000 TaiwSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)30/30SMD = −0.71 (−1.23, −0.19)Likely (favouring treatment)Moderate
Psychological therapies only versus standard care or waitlist
CBT
RMnpta 01.011Depression above thresholdCooper2003 UK
Honey2002 UK
Prendergast2001 Aus
SIGN: 1+NoneRelevantNone83/94RR = 0.78 (0.58, 1.03)Likely (favouring treatment)Moderate
RMnpta 01.012Mean endpoint EPDS scoresCooper2003 UK
Honey2002 UK
Prendergast2001 Aus
SIGN: 1+Significant heterogeneity: random effects model used (−1)RelevantSparse/imprecise data (−1)82/92WMD (random effects) = − 0.99 (−3.31, 1.33)InconclusiveLow
Psychotherapy
RMnpta 01.011Depression above thresholdCooper2003 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)50/52RR = 0.59 (0.38, 0.9)Likely (favouring treatment)Moderate
RMnpta 01.012Mean endpoint EPDS scoresCooper2003 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)45/50WMD = −2.4 (−4.21, −0.59)Likely (favouring treatment)Moderate
IPT
RMnpta 01.011Depression above thresholdO'Hara2000 USSIGN: 1+N/ARelevantNone60/60RR = 0.41 (0.27, 0.62)Very likely (favouring treatment)Moderate
RMnpta 01.012Mean endpoint EPDS scoresO'Hara2000 USSIGN: 1+N/ARelevantNone48/51WMD = −8.5 (−11.25, − 5.75)Very likely (favouring treatment)Moderate
All therapies versus standard care/waitlist - depression symptoms (continuous) at first available follow-up
Psychological therapies at follow-up
RMnpta 03.01Mean EPDSCooper2003 UK
Honey2002 UK
Prendergast2001 Aus
SIGN: 1+N/ARelevantNone81/88WMD = − 0.92 (−2.49, 0.65)UnlikelyHigh
RMnpta 03.02Mean BDIMilgrom2005 AusSIGN: 1+N/ARelevantSparse/imprecise data (−1)12/8WMD = − 8.96 (−17.32, − 0.6)Likely (favouring treatment)Moderate
RMnpta 04.01Depression above thresholdCooper2003 UK
Prendergast2001 Aus
Honey2002 UK
SIGN: 1+N/ARelevantNone90/94RR = 0.73 (0.49, 1.08)Very likely (favouring treatment)Moderate
Acceptability/tolerability
All therapies versus standard care/waitlist - acceptability
RMnpta 05.01All interventions versus standard care or waitlist (all)Cooper2003 UK
Honey2002 UK
Milgrom2005 Aus
O'Hara2000 US
Wickberg1996B Swe
Chen2000 Taiw
SIGN: 1+N/ARelevantNone345/241RR (random effects) = 2.25 (1, 5.04)17%7%N/AModerate
By therapy
RMnpta 05.02Psychological therapiesCooper2003 UK
Honey2002 UK
Milgrom2005 Aus
O'Hara2000 US
SIGN: 1+N/ARelevantNone239/187RR (random effects) = 2.66 (0.84, 8.44)19%7%N/AModerate
RMnpta 05.02Listening visitsCooper2003 UK
Wickberg1996B Swe
SIGN: 1+N/ARelevantNone72/76RR (random effects) = 1.49 (0.6, 3.7)14%9%N/AModerate
RMnpta 05.02Social supportChen2000 TaiwSIGN: 1+N/ARelevantNone34/30RR = 7.97 (0.45, 142.21)12%0%N/AModerate
Dataset by diagnosis
RMnpta 05.03DSM or similarCooper2003 UK
Milgrom2005 Aus
O'Hara2000 US
SIGN: 1+N/ARelevantNone264/165RR (random effects) = 2.24 (0.66, 7.52)18%8%N/AModerate
RMnpta 05.03EPDS >= 12, BDI >= 10 or similarChen2000 Taiw
Honey2002 UK
Wickberg1996B Swe
SIGN: 1+N/ARelevantNone81/76RR = 3.13 (1.06, 9.25)15%4%N/AModerate
Psychological therapies
RMnpta 05.04CBTCooper2003 UK
Honey2002 UK
SIGN: 1+N/ARelevantNone83/94RR = 1.6 (0.48, 5.36)7%4%N/AModerate
RMnpta 05.04PsychotherapyCooper2003 UKSIGN: 1+N/ARelevantNone50/52RR = 2.6 (0.87, 7.75)20%8%N/AModerate
RMnpta 05.04IPTO'Hara2000 USSIGN: 1+N/ARelevantNone60/60RR = 1.33 (0.61, 2.93)20%15%N/AModerate
Therapies versus other therapies
IPT versus psychoeducation (mother only) (antenatal intervention)
RMnpta 019.01Leaving study without attending any sessionsSpinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)25/25RR = 0.5 (0.17, 1.45)16%32%N/AModerate
RMnpta 019.02HDRS >6 by endpoint (ITT)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)25/25RR = 0.84 (0.71, 1)UnlikelyModerate
RMnpta 019.02CGI >2 by endpoint (ITT)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)25/25RR = 0.55 (0.35, 0.84)Very likely (favouring IPT)Moderate
RMnpta 019.03EPDS by endpoint (ITT)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)25/25RR = 0.78 (0.6, 1.03)Likely (favouring IPT)Moderate
RMnpta 019.03BDI by endpoint (ITT)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)25/25RR = 0.67 (0.45, 0.98)Likely (favouring IPT)Moderate
RMnpta 019.04HDRS by endpoint (ITT)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)25/25RR = 0.7 (0.47, 1.04)Likely (favouring IPT)Moderate
RMnpta 019.05BDI (mean)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/17WMD = − 2.15 (−6.27, 1.97)InconclusiveLow
RMnpta 019.05EPDS (mean)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)20/18WMD = − 1.94 (−5.93, 2.05)InconclusiveLow
RMnpta 019.05HRSD-21 (mean)Spinelli2003 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)21/17WMD = − 4.52 (−9.14, 0.1)Likely (favouring IPT)Moderate
Counselling (single session) versus counselling (6 sessions)
RMnpta 020.01Leaving study for any reasonAppleby1997 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)23/21RR = 0.91 (0.35, 2.4)26%29%InconclusiveLow
RMnpta 020.02EPDS at endpoint (3 months postpartum; ITT)Appleby1997 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)23/21WMD = 0.8 (−3.59, 5.19)InconclusiveLow
RMnpta 020.02HRSD at endpoint (3 months postpartum; ITT)Appleby1997 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)23/21WMD = 3.2 (−1.57, 7.97)Likely (favouring 6-session counselling)Moderate
Group exercise versus social support
RMnpta 021.01Leaving study early for any reason before endpointArmstrong2004 AusSIGN: 1+N/ARelevantSparse/imprecise data (−1)12/12RR = 1.5 (0.3, 7.43)25%17%InconclusiveLow
RMnpta 021.02EPDS at endpoint (completer)Armstrong2004 AusSIGN: 1+N/ARelevantSparse/imprecise data (−1)9/10WMD = −7 (−12.32, −1.68)Likely (favouring group exercise)Moderate
Psychoeducation (with partner) versus psychoeducation (mother only)
RMnpta 022.02Major depression by MINI clinicial interview at endpoint (ITT)Misri2000 CanSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/13RR = 0.51 (0.22, 1.18)Likely (favouring psycho- education with partner)Moderate
RMnpta 022.02Major depression by MINI clinicial interview (1 month post-treatment follow-up; ITT)Misri2000 CanSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/13RR = 0.3 (0.1, 0.92)Likely (favouring psycho- education with partner)Moderate
RMnpta 022.03EPDS at endpoint (ITT)Misri2000 CanSIGN: 1+N/ARelevantSparse/imprecise data (−1)15/13WMD = −3.2 (−8.22, 1.82)Likely (favouring psycho- education with partner)Moderate
RMnpta 022.03EPDS at 1-month post- treatment follow-up (ITT)Misri2000 CanSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/13WMD = −6.1 (−10.77, − 1.43)Likely (favouring psycho- education with partner)Moderate
Infant massage (group) versus social support (group)
RMnpta 023.01Leaving study early for any reason before endpointOnozawa2001 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)19/15RR = 2.76 (0.67, 11.41)37%13%InconclusiveLow
RMnpta 023.02EPDS endpoint mean (completer)Onozawa2001 UKSIGN: 1+N/ARelevantSparse/imprecise data (−1)12/13WMD = −4.7 (−7.63, −1.77)Likely (favouring infant massage)Moderate
Acupuncture for depression versus non-depression acupuncture (antenatal intervention)
RMnpta 024.01Leaving study early for any reason before end of acute phaseManber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)20/21RR = 2.1 (0.43, 10.23)20%10%InconclusiveLow
RMnpta 024.02Meeting DSM-IV major depressive disorder criteria at acute phase endpoint (ITT)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)20/21RR = 0.9 (0.37, 2.22)InconclusiveLow
RMnpta 024.03HRDS at acute phase endpoint (completer)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/19WMD = −3 (−8.1, 2.1)InconclusiveLow
RMnpta 024.04BDI at acute phase endpoint (completer)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/19WMD = −3 (−6.85, 0.85)Likely (favouring acupuncuture for depression)Moderate
RMnpta 024.05Not meeting "response" criteria at acute phase endpoint (ITT)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)20/21RR = 0.79 (0.43, 1.45)InconclusiveLow
Acupuncture for depression versus massage (antenatal intervention)
RMnpta 025.01Leaving study early for any reason before end of acute phaseManber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)20/20RR = 4 (0.49, 32.72)20%5%InconclusiveLow
RMnpta 025.02Meeting DSM-IV major depressive disorder criteria at acute phase endpoint (ITT)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)20/20RR = 1.2 (0.44, 3.3)InconclusiveLow
RMnpta 025.03HRDS at acute phase endpoint (completer)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/19WMD = −0.7 (−5.28, 3.88)InconclusiveLow
RMnpta 025.04BDI at acute phase endpoint (completer)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)16/19WMD = −0.8 (−4.29, 2.69)InconclusiveLow
RMnpta 025.05Not meeting "response" criteria at acute phase endpoint (ITT)Manber2004 USSIGN: 1+N/ARelevantSparse/imprecise data (−1)20/20RR = 0.64 (0.37, 1.13)InconclusiveLow
Group counselling versus standard care
RMnpta 026.01BDI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)34/18WMD = − 0.75 (−6.41, 4.91)InconclusiveVery low
RMnpta 026.01BDI (12-month follow-up; completer)Milgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)13/8WMD = − 7.36 (−15.93, 1.21)Likely (favouring group counselling)Low
RMnpta 026.02BAI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)34/18WMD = −2.9 (−8.64, 2.84)InconclusiveLow
Individual counselling versus standard care
RMnpta 027.01BDI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)38/18WMD = − 6.17 (−10.97, − 1.37)Likely (favouring individual counselling)Low
RMnpta 027.01BDI (12-month follow-up; completer)Milgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)24/8WMD = − 11.34 (−18.33, −4.35)Very likely (favouring individual counselling)Low
RMnpta 027.02BAI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)38/18WMD = − 5.85 (−11.81, 0.11)Likely (favouring individual counselling)Low
Counselling (individual) versus counselling (group)
RMnpta 028.01BDI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)38/34WMD = − 5.42 (−10.35, − 0.49)Likely (favouring individual counselling)Low
RMnpta 028.01BDI (12-month follow-up; completer)Milgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)24/13WMD = − 3.98 (−9.96, 2)InconclusiveLow
RMnpta 028.02BAI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)38/34WMD = − 2.95 (−8.25, 2.35)InconclusiveLow
CBT versus counselling (individual)
RMnpta 029.01BDI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)31/38WMD = 1.87 (−2.28, 6.02)InconclusiveLow
RMnpta 029.01BDI (12-month follow-up; completer)Milgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)12/24WMD = 2.38 (−3.28, 8.04)InconclusiveLow
RMnpta 029.02BAI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)31/38WMD = 2.56 (−3.74, 8.86)InconclusiveLow
CBT versus counselling (group)
RMnpta 030.01BDI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)31/34WMD = − 3.55 (−8.68, 1.58)InconclusiveLow
RMnpta 030.01BDI (12-month follow-up; completer)Milgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)12/13WMD = −1.6 (−9.13, 5.93)InconclusiveLow
RMnpta 030.02BAI endpoint (12-week) completerMilgrom2005 AusSIGN: 1− (−1)N/ARelevantSparse/imprecise data (−1)31/34WMD = − 0.39 (−6.47, 5.69)InconclusiveLow

From: APPENDIX 19, EVIDENCE PROFILES

Cover of Antenatal and Postnatal Mental Health
Antenatal and Postnatal Mental Health: The NICE Guideline on Clinical Management and Service Guidance.
NICE Clinical Guidelines, No. 45.
National Collaborating Centre for Mental Health (UK).
Leicester (UK): British Psychological Society; 2007.
Copyright © 2007, The British Psychological Society & The Royal College of Psychiatrists.

All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Enquiries in this regard should be directed to the British Psychological Society.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.