shanghai archives of psychiatry

Transcription

ISSN 1002-0829
CN 31-1564/R
Shanghai Archives of Psychiatry, 2016, Vol. 28, No. 2
•1•
SYSTEMATIC REVIEW AND META-ANALYSIS
Huperzine A for treatment of cognitive impairment in major depressive disorder: a systematic review of
randomized controlled trials
ORIGINAL RESEARCH ARTICLES
A community-based controlled trial of a comprehensive psychological intervention for community residents
with diabetes or hypertension
Disability, psychiatric symptoms, and quality of life in infertile women: a cross-sectional study in Turkey
Clinical investigation of speech signal features among patients with schizophrenia
FORUM
Is the DSM-5 hoarding disorder diagnosis valid in China?
CASE REPORTS
Behavioral and emotional manifestations in a child with Prader-Willi syndrome
Treatment resistant depression or dementia: a case report
BIOSTATISTICS IN PSYCHIATRY (32)
Correlation and agreement: overview and clarification of competing concepts and measures
2016
Vol.28 No. 2
Shanghai Mental Health Center
SHANGHAI ARCHIVES OF PSYCHIATRY
Editorial Advisors
Niufan GU (顾牛范)
Wenyuan WU (吴文源)
Taoyuan XU (徐韬园)
Heqin YAN (严和骎)
Tongji University, Shanghai, China
Fudan University, Shanghai, China
Shanghai Mental Health Center, Shanghai, China
Mingdao ZHANG (张明岛)
Wenwei YAN (颜文伟)
Shanghai Jiao Tong University, Shanghai, China
Zhanpei ZHENG (郑瞻培)
Honorary Editors
Zucheng WANG (王祖承)
Mingyuan ZHANG (张明园)
Editors-in-Chief
Kaida JIANG (江开达)
Michael R. PHILLIPS (费立鹏)
Emory University, Georgia, USA
Managing Editor
Liwei WANG (王立伟)
Associate Editors
John COOPER
Lingjiang LI (李凌江)
Norman SARTORIUS
Xueli SUN (孙学礼)
University of Nottingham, Nottingham, UK
Central South University, Hunan, China
Association for the Improvement of Mental Health Programmes (AMH),
Geneva, Switzerland
Sichuan University, Sichuan, China
Xin YU (于欣)
Peking University, Beijing, China
Biostatistical Editors
Hua HE (贺华)
Ying LU (陆盈)
University of Rochester, New York, USA
Stanford University, California, USA
Xin M. TU (屠心铭)
Systematic Review and Meta-analysis Editor
Chunbo LI (李春波)
Research Methods in Psychiatry Editor
Hui G. CHENG (程辉 )
Assistant Editor
Marlys A. BUEBER (毕曼丽)
Editorial Staff
Bing CAI (蔡冰)
Meng LIU (刘萌)
Tiehong WANG (王铁红 )
Hongxia ZHANG (张红霞)
Wenxia ZHANG (张文霞)
Fei DENG (邓斐)
Yingzhi LIU (刘颖芝)
Manfei XU (徐曼菲)
Jinyi ZHANG (张锦漪)
Editorial Board (in alphabetical order)
Clive E. ADAMS
Paul E. BEBBINGTON
José M. BERTOLOTE
Eric D. CAINE
Joseph R. CALABRESE
William CARPENTER
Raymond CHAN (陈楚侨)
Wei CHEN (陈炜)
Helen Fung-Kum CHIU (赵凤琴)
Joseph COYLE
Joseph F. CUBELLS
John DAVIS
Diego DE LEO
Yasong DU (杜亚松)
University of Nottingham, Nottingham, UK
São Paulo State University (UNESP), São Paulo, Brazil
Case Western Reserve University, Ohio, USA
Chinese Academy of Sciences, Beijing, China
Chinese University of Hong Kong, Hong Kong, China
Griffith University, Queensland, Australia
University College London, London, UK
University of Maryland, Maryland, USA
Zhejiang University, Zhejiang, China
Harvard University, Massachussetts, USA
University of Illinois at Chicago, Illinois, USA
Editorial Board (in alphabetical order, continued)
Naihua DUAN
Columbia University, New York, USA
David L. DUNNER
Center for Anxiety and Depression, Washington, USA
Xiaoduo FAN
Yiru FANG (方贻儒)
Jonathan FLINT
Sophia FRANGOU
Robert FREEDMAN
David GOLDBERG
Byron J. GOOD
Kyooseob HA
Mohamad Hussain Bin HABIL
Wei HAO (郝伟)
Yanling HE (何燕玲)
Helen HERRMAN
Teh-wei HU
Rachel JENKINS
Jianlin JI (季建林)
Fujun JIA (贾福军)
Tianzi JIANG (蒋田仔)
Hua JIN
Shigenobu KANBA
Kenneth S. KENDLER
Ronald C. KESSLER
Murad M. KHAN
Arthur KLEINMAN
John H. KRYSTAL
Ee Heok KUA
Huafang LI (李华芳)
Huichun LI (李惠春)
Tao LI (李涛)
Xiaobai LI (李晓白)
Jeffrey A. LIEBERMAN
Walter LING
Tiebang LIU (刘铁榜)
Lin LU (陆林)
Zheng LU (陆峥)
Mario MAJ
Jair de Jesus MARI
Graham MELLSOP
Fumitaka NODA
Vikram PATEL
Mark H. RAPAPORT
Darrel A. REGIER
Shekhar SAXENA
Shenxun SHI (施慎逊)
Morton SILVERMAN
Christopher Paul SZABO
Phern-Chern TOR
Peter TYRER
Pichet UDOMRATN
Chuanyue WANG (王传跃)
Gaohua WANG (王高华)
Yuanjia WANG
Shifu XIAO (肖世富)
Zeping XIAO (肖泽萍)
Bin XIE (谢斌)
Yifeng XU (徐一峰)
Yanchun YANG (杨彦春)
Albert YEUNG
Yufeng ZANG (臧玉峰)
Dai ZHANG (张岱)
Ning ZHANG (张宁)
Yalin ZHANG (张亚林)
Zhijun ZHANG (张志珺)
Jingping ZHAO (赵靖平)
Min ZHAO (赵敏)
Xudong ZHAO (赵旭东)
Dongfeng ZHOU (周东丰)
Ziqing ZHU (朱紫青)
Douglas ZIEDONIS
University of Massachusetts, Massachusetts, USA
Oxford University, Oxford, UK
University of Colorado, Colorado, USA
University of Malaya, Kuala Lumpur, Malaysia
University of California, Berkeley, California, USA
Chinese Academy of Sciences, Beijing, China
Kyushu University, Fukuoka, Japan
National University of Singapore, Singapore
Zhejiang University, Zhejiang, China
China Medical University, Beijing, China
University of California, Los Angeles, California, USA
University of Naples, Naples, Italy
University of Auckland, Auckland, New Zealand
Public Health Foundation of India, New Delhi, India
Uniformed Services University, USA
University of the Witwatersrand, Johannesburg, South Africa
Imperial College London, London, UK
Capital Medical University, Beijing, China
Shanghai Health Bureau, Shanghai, China
Tongji University, Shanghai, China
Center for Disease Control and Prevention, Shanghai, China
King’s College London, London, UK
Institute of Psychiatry, London, UK
Seoul National University, Seoul, South Korea
University of Melbourne, Victoria, Australia
Kings College London, London, UK
Guangdong Mental Health Center, Guangdong, China
University of California, San Diego, California, USA
Virginia Commonwealth University, Virgina, USA
Aga Khan University, Karachi, Pakistan
Yale University, Connecticut, USA
Shenzhen Psychiatry Research Center, Guangdong, China
Shanghai Tongji University, Shanghai, China
Universidade Federal de São Paulo, São Paulo, Brazil
Mejiro SOLA Clinic, Tokyo, Japan
World Health Organization, Geneva, Switzerland
University of Colorado, Colorado, USA
National University of Singapore, Singapore
Prince of Songkla University, Songkla, Thailand
Wuhan University, Hubei, China
Hangzhou Normal University, Zhejiang, China
Nanjing Medical University, Jiangsu, China
Southeast University, Jiangsu, China
University of Massachusetts, Massachusetts, USA
Free e-subscription
to SHARP
Publish your work in both
English and Chinese!
SHARP
Shanghai Archives of Psychiatry Since 1959
A Bridge Connecting Chinese and International Psychiatry
Why register for a free e-subscription
to SHARP?
- Cutting-edge psychiatric research from mainland
China
- Up-to-date information on the rapid evolution of
China’s mental health delivery system
- Forums that present Chinese and international
opinions about controversial issues related to mental
health
- Systematic reviews and meta-analyses that integrate
data from both Chinese and international sources
- A unique Biostatistics in Psychiatry section
Register for free e-subscription at:
www.shanghaiarchivesofpsychiatry.org/subscription.html
OR
Scan the following QR code:
Why publish in SHARP?
- The only psychiatric journal from mainland China
published in both English and Chinese
- An opportunity for international authors to have
their work read in Chinese by the 20,000 mental
health professionals in China
- An open access journal with free downloads of full
text articles on PubMed Central
- Extensive technical support from the editorial office
and from the distinguished editorial board that
includes over 100 prominent Chinese and international experts
- Articles are indexed in PubMed, PsycINFO, EBSCO,
SCOPUS and 12 other international and Chinese
databases
- The table of contents is published in the American
Journal of Psychiatry
- No page charges or other fees
See the Instructions for Authors at:
http://www.shanghaiarchivesofpsychiatry.org/en/instrution-forauthors.html
Volume 28 • Number 2 • April 2016
61 In this issue (April 2016)
Forum
Systematic review and meta-analysis
103 Is the DSM-5 hoarding disorder diagnosis
valid in China?
64 Huperzine A for treatment of cognitive
impairment in major depressive
disorder: a systematic review of
randomized controlled trials
Case Reports
Wei ZHENG, Ying-Qiang XIANG,
Gabor S. UNGVARI, Helen F.K. CHIU,
Chee H. NG, Ying WANG, Yu-Tao XIANG
Original research articles
72 A community-based controlled trial
of a comprehensive psychological
intervention for community residents
with diabetes or hypertension
Qingzhi ZENG, Yanling HE, Zhenyu SHI,
Weiqing LIU, Hua TAO, Shiming BU,
Donglei MIAO, Ping LIU, Xuanzhao ZHANG,
Xiaoping LI, Xuejun QI, Qin ZHOU
86 Disability, psychiatric symptoms, and
quality of life in infertile women: a
cross-sectional study in Turkey
Hacer SEZGIN, Cicek HOCAOGLU,
Emine Seda GUVENDAG-GUVEN
95 Clinical investigation of speech
signal features among patients with
schizophrenia
Jing ZHANG, Zhongde PAN, Chao GUI,
Jie ZHU, Donghong CUI
Bimonthly
Founded in 1959
April, 2016
ISSN 1002-0829
CN 31-1564/R
Copyright© 2016 by Editorial Department of the Shanghai
Archives of Psychiatry
Responsible Institute: Shanghai Mental Health Center
Sponsor: Shanghai Municipal Commission of Health and
Family Planning
All articles published represent the opinions of the authors; they do
not reflect the official policy of the Editorial Board unless this is clearly
specified.
Zhen WANG, Yuan WANG, Qing ZHAO,
Kaida JIANG
106 Behavioral and emotional manifestations
in a child with Prader-Willi syndrome
Satyakam MOHAPATRA, Udit Kumar PANDA
109 Treatment resistant depression or
dementia: a case report
Zhongyong SHI, Shifu XIAO, Xia LI
Biostatistics in psychiatry (32)
115 Correlation and agreement: overview and
clarification of competing concepts and
measures
Jinyuan LIU, Wan TANG, Guanqin CHEN, Yin LU,
Changyong FENG, Xin M. TU
A1 Contents of the American Journal of Psychiatry
(February 2016)
A2 Contents of the American Journal of Psychiatry
(March 2016)
Editing: Editorial Board of the Shanghai Archives of Psychiatry
Editors-in-Chief: Kaida JIANG, Michael R. PHILLIPS
Publisher: Shanghai Municipal Bureau of Publishing
Address: 600 Wan Ping Nan Lu, 200030 Shanghai, China
Telephone number: +86 (021) 34773296
Website: www.shanghaiarchivesofpsychiatry.org
Email: [email protected]
Printing: Shanghai Mier Printing Co., Ltd.
Domestic Subscription: Shanghai Magazine Subscription Office
Publication format: Open Access
Chinese publication number: 4-798
Advertisement permit number: 3100420080042
Single issue price: 9.00 RMB
• 61 •
•In this issue (April 2016)•
This issue begins with a systematic review and metaanalysis by Zheng and colleagues[1] about the use of a
traditional Chinese medicine – Huperzine A (HupA) – as
an adjunctive treatment for depression. The rationale
for this treatment is that acetylcholinesterase (AChE)
inhibitors may reduce the cognitive impairment that
often accompanies depressive episodes and HupA
is a powerful AChE inhibitor. After an exhaustive
literature search in English language and Chinese
language journals, the authors only found three
randomized controlled trials (with a pooled sample
of 238 individuals) comparing monotherapy with
an antidepressant to combined treatment with an
antidepressant and HupA. When pooling results, there
was no significant difference between groups in the
degree of improvement in depressive symptoms, but
there was significantly greater improvement in cognitive
functioning in the group that received adjunctive HupA
(as assessed by the Wisconsin Card Sorting Test and
the Wechsler Memory Scale-Revised). However, the
three studies were open label (i.e., non-blinded) and
only followed subjects for a mean of 6.7 weeks, so the
studies were classified as ‘low-quality’. Thus, more
rigorously conducted studies that follow participants
longer are needed to confirm this important result. This
is an example of a common problem in using Traditional
Chinese Medicine (TCM): the results are often
promising, but the lack of rigorous scientific proof limits
the acceptance of the results in non-Chinese settings.
The first original research article by Zeng and
colleagues [2] reports on a large community-based
intervention aimed at reducing the severity of depressive
and anxiety symptoms in community residents receiving
treatment for diabetes or hypertension. China, like
other low- and middle-income countries, does not
have sufficient psychiatric manpower to provide
individualized treatment to persons with chronic
illnesses who have comorbid depression or anxiety, so
the authors adapted the community-based IMPACT
model developed in the United States [3] for use in
Shanghai. This approach includes community-based
health education about psychological problems, peergroup support for persons with mild depression or
anxiety, and individual counseling (using the Problem
Solving Treatment for Primary Care [4] method) for
those with moderate or severe depression or anxiety.
Baseline evaluations and 6-month follow-up evaluations
using self-completion instruments assessing depressive
symptoms, anxiety symptoms, and quality of life were
completed by 3039 individuals in the intervention
group and 1239 individuals in the treatment as usual
group (i.e., standard follow-up care of chronic physical
illnesses). All community members in the intervention
communities were exposed to the health education
initiative, but participation of eligible individuals in the
peer-support groups was low (31%) and participation
of eligible individuals in the individual counseling
was very low (9%). Nevertheless, after 6 months
the improvement in depressive symptoms, anxiety
symptoms, and quality of live was significantly greater
in the intervention group than in the control group. This
study demonstrates the feasibility of such communitybased interventions for decreasing the severity of comorbid psychological symptoms in persons with chronic
physical illnesses, but further work is needed to increase
the participation rates in the support services provided
for persons with mild and moderate depression and
anxiety.
The second original research article by Sezgin and
colleagues[5] is a cross-sectional study from urban Turkey
that compares self-reported psychological symptoms
and disability between 100 married women seeking
treatment for infertility and 100 fertile married women.
The authors used Turkish versions of the Hospital Anxiety
and Depression Scale (HADS),[6] the Brief Disability
Questionnaire (BDQ),[7] and the Short Form Health
Survey (SF-36)[8] to compare the self-reported levels of
depressive symptoms, anxiety symptoms, disability, and
quality of life of the two groups of respondents. The
study found no significant difference in the self-reported
levels of depressive or anxiety symptoms, but the
respondents in the infertile group reported significantly
greater disability, and a significantly lower quality of life.
Thus western assumptions about the close relationship
between social stressors, psychological symptoms, and
functioning may not hold true in non-western countries
or for specific types of stressors (such as infertility).
But this was a relatively small cross-sectional study;
larger, longitudinal studies are needed to confirm these
interesting results.
The third original research article by Zhang and
colleagues[9] considers the possibility of using easily
obtained acoustic features of speech (i.e., ‘speech
signal features’), which can reflect the emotional
responsiveness of the speaker, as biomarkers for
schizophrenia. The authors analyzed 10 acoustic
features of a 15-minute speech sample obtained by
smart phone from 26 inpatients with schizophrenia
and compared them to the features of speech samples
obtained from 30 healthy controls. They also assessed
the severity of the patients’ symptoms at baseline and
obtained a second speech sample from the patients one
week later. The ten speech signal features (6 prosody
features, formant bandwidth and amplitude, and two
spectral features) were stable over time (intraclass
correlation coefficients ranging from 0.55 to 0.88), but
only two of the features (the two spectral features) were
significantly different between patients and controls.
• 62 •
There were significant correlations between some of
the speech features and the severity of the negative
symptoms of schizophrenia. These finding provide some
support for the potential value of acoustic features of
speech as biomarkers for schizophrenia, particularly
the negative symptoms of schizophrenia. But larger
studies that monitor the acoustic features over time
as patients’ symptoms wax and wane are needed to
determine whether or not these features can accurately
differentiate persons with and without schizophrenia,
and whether or not they can be used as markers of the
severity of the illness.
The Forum by Wang and colleagues[10] addresses
a perennial issue: whether or not the diagnostic
criteria for a condition described in the 5th edition of
the American Psychiatric Association’s Diagnostic and
Statistical Manual (DSM-5)[11] are culturally appropriate
for China. China had previously developed its own
psychiatric classification system –CCMD3[12]—but this
has now been abandoned; in clinical settings the official
recommendation from the government is to use the
classification system of the World Health Organization
(ICD-10[13]), but most clinical researchers prefer to use
the DSM system. However, for certain disorders there
are serious concerns about the validity of diagnostic
criteria developed for use in the American population in
other cultural settings. In this particular case the authors
discuss hoarding disorder which has been ‘upgraded’
from one of the symptoms of obsessive-compulsive
disorder in the 4th edition of the DSM (DSM-IV)[14] to
a separate disorder under the ‘Obsessive-Compulsive
and Related Disorders’ chapter of DSM-5. The rationale
for this change was that research that was primarily
conducted in the United States and other Western
countries had found distinct differences between the
clinical symptoms, family history, and neuroimaging
characteristics of individuals with pathological hoarding
and those with obsessive compulsive disorder in the
absence of hoarding. After review of available literature
from China and other East Asian countries, the authors
conclude that pathological hoarding is relatively
common in East Asia and that the DSM-5 classification
of this as a separate disorder is justified in East Asia.
However, they caution that in countries like China with
a recent history of material scarcity, thriftiness is often
a culturally sanctioned trait, so the hoarding behavior
needs to be associated with significant distress and
with substantial social impairment before it should be
considered a psychiatric diagnosis.
The first case report from India by Satyakam
and Panda[15] is about a 9-year-old girl with PraderWilli syndrome who was brought to a psychiatric
hospital by her family because of serious behavioral
problems including irritability, emotional lability,
and temper tantrums. The family reported delayed
motor and language development, over-eating, and
unexplained emotional outbursts. On physical exam
she was obese (BMI of 43), had small hands and feet,
almond-shaped palpebral fissures, and self-inflicted
excoriated skin lesions. She had an IQ of 40, but the
computed tomography of her brain was normal.
Diagnosed with Prader-Willi syndrome based on the
clinical presentation, she was treated with low-dose
antipsychotics (risperidone 1mg/d). After 8 weeks of
treatment the behavioral outbursts and self-injurious
behavior improved significantly. In low- and middleincome countries without the resources to conduct
sophisticated genetic testing, the diagnosis of such
rare conditions depends on the correct identification
of the typical clinical symptoms; given the unfamiliarity
of most clinicians with such conditions, it is likely that
many of them remain undiagnosed and untreated.
The second case report by Shi and colleagues [16]
discusses an increasingly common dilemma in China as
the population ages: differentiating chronic, treatmentresistant depression from the early onset of dementia.
In this case of a 78-year-old woman with previous
episodes of major depression, the clinical picture was
complicated by her long-term use of a reserpine-based
hypertensive. She presented with typical symptoms
of both depression and dementia; after 8 weeks of
inpatient treatment (including changing her antihypertensive medication) the depressive symptoms
improved but the cognitive symptoms did not. She
subsequently developed cancer at which point the
depressive symptoms exacerbated. The authors
conclude that in such complicated cases of elderly
patients with symptoms of both depression and
dementia it will often be necessary to follow the course
of the symptoms for one or two years before it can
be determined whether the cognitive symptoms are
secondary to depression or a newly emerging dementia
(or both).
The Biostatistics in Psychiatry paper by Liu and
colleagues[17] discusses an important topic that is often
misused by statistically-challenged researchers: the
difference between agreement and correlation. The
degree of agreement between variables is assessed
when considering the relationship between variables
that are different measures of the same construct; the
level of correlation between variables is assessed when
considering the relationship of variables that measure
different constructs. The authors discuss the different
statistics used to evaluate these two measures of
association, emphasize the importance of considering
the distribution of the variables being considered
(continuous or non-continuous), and provide several
examples of the issues than need to be considered
when assessing commonly used measures of association
such as the Pearson correlation coefficient and the
intraclass correlation coefficient.
[Shanghai Arch Psychiatry. 2016; 28(2): 61-63. doi: http://dx.doi.org/10.11919/j.issn.1002-0829.216050]
• 63 •
References
1.
2.
3.
Zheng W, Xiang YQ, Ungvari GS, Chiu HFK, Ng CH, Wang Y,
et al. Huperzine A for treatment of cognitive impairment
in major depressive disorder: a systematic review of
randomized controlled trials. Shanghai Arch Psychiatry.
2016; 28(2): 64-71. doi: http://dx.doi.org/10.11919/
j.issn.1002-0829.216003
Zeng QZ, He YL, Shi ZY, Liu WQ, Tao H, Bu SM, et al. A
community-based controlled trial of a comprehensive
psychological intervention for community residents with
diabetes or hypertension. Shanghai Arch Psychiatry.
j.issn.1002-0829.216016
Katon W, Unutzer J, Wells K, Jones L. Collaborative
depression care: history, evolution and ways to enhance
dissemination and sustainability. Gen Hos Psychiatry.
j.genhosppsych.2010.04.001
4.
Hegel M, Areán P. Problem-solving Treatment for Primary
Care: A Treatment Manual for Project Impact. (Thesis
dissertation). Dartmouth University; 2003
5.
Sezgin H, Hocaoglu Cicek, Guvendag-Guven ES. Disability,
psychiatric symptoms, and quality of life in infertile women:
a cross-sectional study in Turkey. Shanghai Arch Psychiatry.
j.issn.1002-0829.216014
6.
Aydemir O, Guvenir T, Kuey L, Kultur S. [Reliability and
validity of the Turkish version of the Hospital Anxiety and
Depression Scale]. Turk Psikiyatri Derg. 1997; 8(3): 280-287.
Turkish
7.
Kaplan I. [The relationship between mental disorders and
disability in patients admitted to the semi-rural health
centers]. Turk Psikiyatri Derg. 1995; 6(2): 169-179. Turkish
8.
Koçyigit H, Aydemir O, Fisek G, Olmez N, Memis A. [The
reliability and validity of the Turkish version of Short Form36 (SF-36)]. İlaç ve Tedavi Dergisi. 1999; 12(3): 102-106.
Turkish
9.
Zhang J, Pan ZD, Gui C, Zhu J, Cui DH. Clinical investigation of
speech signal features among patients with schizophrenia.
Shanghai Arch Psychiatry. 2016; 28(2): 95-102. doi: http://
dx.doi.org/10.11919/j.issn.1002-0829.216025
10. Wang Z, Wang Y, Zhao Q, Jiang Kd. Is the DSM-5 hoarding
disorder diagnosis valid in China? Shanghai Arch Psychiatry.
j.issn.1002-0829.215054
11. American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders, Fifth Edition (DSM-5). Arlington
VA: American Psychiatric Association; 2013
12. Chinese Medical Association. [Chinese Mental Disorders
Classification and Diagnostic Criteria, Third Edition (CCMD3)]. Jinan: Shandong Science and Technology Press; 2001.
Chinese
13. World Health Organization. The ICD-10 Classification of
Mental and Behavioural Disorders: Clinical Descriptions and
Diagnostic Guidelines. Geneva: World Health Organization;
1992
14. American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders, Fourth Edition (DSM-IV).
Washington, DC: American Psychiatric Association; 1990
15. Satyakam M, Panda UK. Behavioral and emotional
manifestations in a child with Prader-Willi syndrome.
Shanghai Arch Psychiatry. 2016; 28(2): 106-109. doi: http://
dx.doi.org/10.11919/j.issn.1002-0829.215110
16. Shi ZY, Xiao SF, Li X. Treatment resistant depression or
dementia: a case report. Shanghai Arch Psychiatry.
2016; 28(2): 109-114. doi:http://dx.doi.org/10.11919/
j.issn.1002-0829.215085
17. Liu JY, Tang W, Chen GQ, Lu Y, Feng CY, Tu XM. Correlation
and agreement: overview and clarification of competing
concepts and measures. Shanghai Arch Psychiatry.
j.issn.1002-0829.216045
• 64 •
•Systematic review and meta-analysis•
Huperzine A for treatment of cognitive impairment in major
depressive disorder: a systematic review of randomized
controlled trials
Wei ZHENG1, Ying-Qiang XIANG2, 3, Gabor S. UNGVARI4, 5, Helen F.K. CHIU6, Chee H. NG7,
Ying WANG8, Yu-Tao XIANG9,*
Background: Acetylcholinesterase (AChE) inhibitors have been shown to be effective in treating cognitive
impairment in animal models and in human subjects with major depressive disorder (MDD). Huperzine A
(HupA), a Traditional Chinese Medicine derived from a genus of clubmosses known as Huperzineserrata, is
a powerful AChE inhibitor that has been used as an adjunctive treatment for MDD, but no meta-analysis on
HupA augmentation for MDD has yet been reported.
Aim: Conduct a systematic review and meta-analysis of randomized controlled trials (RCTS) about HupA
augmentation in the treatment of MDD to evaluate its efficacy and safety.
Methods: Two evaluators independently searched nine English-language and Chinese-language databases,
selected relevant studies that met pre-determined inclusion criteria, extracted data about outcome and
safety, and conducted quality assessments and data synthesis.
Results: Three low-quality RCTs (pooled n=238) from China were identified that compared monotherapy
antidepressant treatment for depression versus combined treatment with antidepressants and HupA.
Participants in the studies ranged from 16 to 60 years of age. The average duration of adjunctive
antidepressant and HupA treatment in the studies was only 6.7 weeks. All three studies were open label
and non-blinded, so their overall quality was judged as poor. Meta-analysis of the pooled sample found no
significant difference in the improvement in depressive symptoms between the two groups (weighted mean
difference: -1.90 (95%CI: -4.23, 0.44), p=0.11). However, the adjunctive HupA group did have significantly
greater improvement than the antidepressant only group in cognitive functioning (as assessed by the
Wisconsin Card Sorting Test and the Wechsler Memory Scale-Revised) and in quality of life. There was no
significant difference in the incidence of adverse drug reactions between groups.
Conclusions: The data available on the effectiveness and safety of adjunctive treatment using HupA in
patients with MDD who are receiving antidepressants is insufficient to arrive at a definitive conclusion
about its efficacy and safety. Pooling of the data from three low-quality RCTs from China found no advantage
of adjunctive HupA in the treatment of depressive symptoms, but adjunctive treatment with HupA was
associated with a faster resolution of the cognitive symptoms that frequently accompany MDD.
Trial registration number: CRD42015024796 (http://www.crd.york.ac.uk/prospero/)
Key words: depression; meta-analysis; cognitive function; huperzine A; adjunctive treatment
1
The Affiliated Brain Hospital of Guangzhou Medical University (Guangzhou Huiai Hospital), Guangzhou, China
China Clinical Research Center for Mental Disorders, Beijing, China, and Center of Depression, Beijing Institute for Brain Disorders, Beijing, China
3
Beijing Key Laboratory of Mental Disorders, Beijing Anding Hospital, Capital Medical University, Beijing, China
4
School of Psychiatry and Clinical Neurosciences, University of Western Australia, Perth, Australia
5
University of Notre Dame Australia / Marian Centre, Perth, Australia
6
Department of Psychiatry, Chinese University of Hong Kong, Hong Kong SAR, China
7
Department of Psychiatry, University of Melbourne, Melbourne, Victoria, Australia
8
Institute of Chinese Medical Sciences, University of Macau, Macao SAR, China
9
Unit of Psychiatry, Faculty of Health Sciences, University of Macau, Macao SAR, China
2
*correspondence: Dr. Yu-Tao Xiang, 3/F, Building E12, Faculty of Health Sciences, University of Macau, Avenida da Universidade, Taipa, Macau SAR,
China. E-mail: [email protected]
A full-text Chinese translation of this article will be available at http://dx.doi.org/10.11919/j.issn.1002-0829.216003 on August 25, 2016.
1. Introduction
Major depressive disorder (MDD) is a common
psychiatric illness that is often associated with cognitive
dysfunction.[1] One hypothesis about the mechanism
of cognitive decline in MDD links it to decreasing
acetylcholinesterase (AChE) activity of the cholinergic
system in the hippocampus, frontal cortex, and
septum.[2] Some studies suggest that AChE inhibitors
(e.g., donepezil,[3] rivastigmine,[4] and galantamine[5])
can ameliorate cognitive impairment in animal models
of depression and in humans with MDD.[5-7] Huperzine
A (HupA) is a Traditional Chinese Medicine (TCM)
isolated from Huperzineserrata (a genus of clubmosses),
also known as ground pines or creeping cedar, in the
family Lycopeodiaceae (a family of fern-allies). It is a
powerful, highly specific, and reversible inhibitor of
AChE.[8-10] Because of its popularity as a TCM medication
in mainland China, extensive clinical experience and
research about HupA in China may help clarify the
mechanism of action for its potential efficacy in the
treatment of MDD. However, to date no systematic
review or meta-analysis on HupA augmentation for
MDD has been published. The primary aim of this study
was to conduct a systematic review and meta-analysis
about the efficacy and safety of HupA in the treatment
of MDD based on published RCTs identified by searching
international and Chinese databases.
2. Methods
2.1 Types of studies
All publications of randomized controlled trials (RCTs)
which reported on the efficacy and/or safety of
antidepressants combined with HupA in the treatment
of MDD were eligible for inclusion. Case reports/series,
observational trials, meta-analyses, and systematic
reviews were excluded.
2.2 Outcome measures
The primary outcome measure of interest was
cognitive function measured by the Wisconsin Card
Sorting Test (WCST)[11] or the Wechsler Memory ScaleRevised, Chinese version (WMS-RC).[12] Key secondary
outcomes were improvement in depressive and anxiety
symptoms assessed by the Hamilton Depression
Rating Scale (HAMD) [13] and the Hamilton Anxiety
Rating Scale (HAMA),[14] self-reported quality of life
assessed by the General Quality of Life Inventory of the
World Health Organization (WHOQOL-100),[15] causes
for discontinuation of treatment, and adverse drug
reactions measured by the Dosage Record Treatment
Emergent Symptom Scale (DOTES).[16] Clinical outcomes
were based on intent-to-treat (ITT) analysis.
2.3 Selection of studies
PubMed, PsycINFO, Embase, Cochrane Library
databases, the Cochrane Controlled Trials Register,
• 65 •
ClinicalTrials.gov (https://www.clinicaltrials.gov/),
and Chinese databases (WanFang Database, Chinese
Biomedical database, and China Journal Net) were
searched from the inception of the databases through
March 12, 2016 using the following search terms:
(Depressive Disorders OR Disorder, Depressive OR
Disorders, Depressive OR Neurosis, Depressive OR
Depressive Neuroses OR Depressive Neurosis OR
Neuroses, Depressive OR Depression, Endogenous OR
Depressions, Endogenous OR Endogenous Depression
OR Endogenous Depressions OR Depressive Syndrome
OR Depressive Syndromes OR Syndrome, Depressive
OR Syndromes, Depressive OR Depression, Neurotic
OR Depressions, Neurotic OR Neurotic Depression OR
Neurotic Depressions OR Melancholia OR Melancholias
OR Unipolar Depression OR Depression, Unipolar OR
Depressions, Unipolar OR Unipolar Depressions) AND
(Huperzine A OR Huperzine OR HupA) AND (randomized
controlled trial OR controlled clinical trial OR randomized
OR placebo OR drug therapy OR randomly OR trial OR
groups). We also hand-searched reference lists from
identified and relevant review articles for additional
studies and contacted authors for unpublished data.
2.4 Data extraction
Two authors (ZW and XYQ) independently conducted
the literature search and extracted the data. Any
disagreement was resolved by a third author (XYT). Data
presented only in graphs and figures were extracted
whenever possible. Authors were contacted to obtain
missing information or clarification if possible. If cases
were from multicenter studies, whenever possible, data
were extracted separately for each center.
2.5 Statistical methods
We used RevMan (version 5.1.7.0) in this meta-analysis
according to the recommendations of the Cochrane
Collaboration. For continuous data, weighted mean
difference (WMD) with 95% CI was used to compare
groups, and for dichotomous data, risk ratio (RR)
with 95% confidence intervals (Cis) were computed
to compare groups. The I2 statistic assessed statistical
heterogeneity between the three studies: when I2≥50%,
a random effects model was used;[17] otherwise, a fixed
effect model was employed.[18] All analyses were twotailed with alpha set at 0.05.
2.6 Risk of bias assessment
The methods of random sequence generation (selection
bias), allocation concealment (selection bias), blinding of
participants and personnel (performance bias), blinding
of outcome assessment (detection bias), incomplete
outcome data (attrition bias), selective reporting
(reporting bias), and other biases were assessed using
the Risk of Bias (ROB) scale developed to assess RCTs by
the Cochrane Collaboration.[19]
• 66 •
3. Results
3.1 Results of the literature search
The search yielded 54 potentially relevant articles, of
which four articles were published in English and 50
in Chinese. Of the 54 studies, 3 RCTs met the inclusion
criteria.[20-22] As shown in Figure 1, the total number of
subjects included in the three studies was 238, with 119
receiving an antidepressant augmentated with HupA
and 119 only receiving an antidepressant.
3.2 The characteristics of included studies
As shown in Table 1, all three RC Ts [20-22] were
conducted in China and used the criteria of the
Chinese Classification of Mental Disorders, 3rd edition
(CCMD-3)[23] to diagnose depression. Males accounted
for 45.4% of the sample (range 30% to 58% in the
three studies), the weighted mean age of participants
was 29.6 (range 16-60) years; and the weighted mean
duration of illness was 3.3 (range 1.2 to 5.2) years. The
weighted mean duration of the treatment trial reported
in the studies was 6.7 (range 6-8) weeks. None of the
studies were supported by pharmaceutical companies.
3.3 Assessment of risk of bias
The risk of different types of biases of the three studies
is shown in Table 2. Two studies [21-22] mentioned
“random” assignment without a description of the
method of randomizing, and one RCT [20] was rated
as high risk of selection bias because patients were
Figure 1. Identification of included studies
54 articles published before May 12, 2016 were identified using a standard search strategy and other sources
(see methods section):
• 24 from China Journal Net
• 14 from WanFang Database
• 12 from Chinese Biomedical database
•
2 from Embase
•
1 from PubMed
•
1 from Cochrane Library databases
•
0 from PsycINFO
•
0 from Cochrane Controlled Trials Register
•
0 from ClinicalTrials.gov (https://www.clinicaltrials.gov/)
6 duplicates removed
48 unduplicated studies; 44 published in Chinese, 4 in English
40 articles excluded based on title and abstract
8 full-text articles assessed for eligibility
5 full-text articles excluded:
• 3 had no major depressive disorder diagnosis
• 1 review
• 1 animal study
3 studies included in qualitative synthesis and in meta-analysis
• 67 •
Table 1. Characteristics of included studies
Diagnosis
Diagnostic Weighted
Male
Trial
criteria mean age
duration Country
in years
n
Mean
(%)
(weeks)
(range)
illness
duration
Design
Study
N
Setting
Interventions:
[M] mean dose (mg/day)
[R] range (mg/day)
[n] number of patients
Outcome
assessments
Open-label
Gao
100 Inpatients
2007[20]
and
outpatients
6
China
MDD
CCMD-3
5.2 years
30.4
(18-50)
30
(30%)
1. FLU(M=NR; R=20-40) + HupA
(fixed dose at 0.3); n=50
2. FLU(M=NR; R=20-40); n=50
HAMD;
WCST;
WHOQOL-100
Yang
78
2010[21]
Open-label
Inpatients
and
outpatients
8
China
MDD
CCMD-3
2.5 years
29.9
(18-60)
45
(58%)
1. FLU(M=NR; R=20-40) + HupA
(fixed dose at 0.3); n=39
2. FLU(M=NR; R=20-40); n=39
HAMD;
WMS-RC
Liu
60
2010[22]
Open-label
Inpatients
and
outpatients
China
MDD
CCMD-3
1.2 years
27.9
(16-48)
33
(55%)
1. VEN(M=107; R=50-150) + HupA HAMD;
(M=NR; R=0.1-0.2); n=30
HAMA;
2. VEN(M=105; R=50-150); n=30
DOTES
6
MDD, Major Depressive Disorder
CCMD-3, Chinese Mental Disorders Classification and Diagnostic
Criteria, Third Edition[23]
FLU, fluoxetine
NR, not recorded
HupA, huperzine A
HAMD, Hamilton Depression Rating Scale[13]
WCST, Wisconsin Card Sorting Test[11]
WHOQOL-100, General Quality of Life Inventory of World Health
Organization[15]
WMS-RC, Wechsler Memory Scale-Revised, Chinese version[12]
VEN, venlafaxine
HAMA, Hamilton Anxiety Rating Scale[14]
DOTES, Dosage Record Treatment Emergent Symptom Scale[16]
Table 2. Evaluation of risk of bias in the three included studies
study
sequence
generation
blinding of
allocation
sequence participants and
personnel
concealment
blinding of
outcome
assessment
incomplete
outcome
data
selective
outcome
reporting
other potential
threats
to validity
Gao
2007[20]
high
high
high
high
low
N/A
low
Yang
2010[21]
N/A
high
high
high
low
N/A
low
Liu
2010[22]
N/A
high
high
high
low
N/A
low
N/A=no information available
classified into two groups according to the order of
admission. None of the studies were blinded so the
risk of allocation bias, performance bias, and detection
bias were high. The studies reported the outcomes of
all enrolled subjects, so the risk of attrition bias was
low; but in the absence of study registration materials it
was impossible to determine whether or not there was
selective reporting (i.e., reporting bias). There was no
evidence of other types of biases (e.g., drug company
sponsorship of the study). Overall, all three studies were
considered at high risk of bias and, thus, relatively lowquality studies.
Because there were only three RCTs included in the
meta-analysis, publication bias could not be tested.[24]
3.4 Changes in severity of depressive symptoms
In all three studies there were differences between
groups in changes of the total HAMD score over
• 68 •
Figure 2. Adjunctive Huperzine A for MDD: forest plot for improvement in depressive symptoms assessed by
change in total score of the Hamilton Depression Scale
Experimental
Study
Total Mean
Gao 2007[20]
Liu 2010
[22]
Yang 2010[21]
Random effects model
Control
SD Total Mean
Mean difference
SD
MD
95%CI
Weight (random)
50 12.06 6.52
50 13.04 7.30
0.98
[ 3.69; 1.73]
32.5%
39
6.11 3.47
39
9.98 5.77
3.87 [ 5.98; 1.76]
39.3%
30
8.10 6.30
30
8.30 6.20
0.20
28.1%
119
119
[ 3.36; 2.96]
1.90 [ 4.23; 0.44]
100
Heterogeneity: I2=57.5%, tau2=2.444, p=0.0948
Test for overall eﬀect: Z=1.59 p=0.1111
20
15
favors experimental
the study period. As shown in Figure 2, one of the
studies[22] reported a significantly greater reduction
of depressive symptoms (based on the HAMD) when
adjunctive HupA was provided to patients with MDD
being treated with antidepressants, but the other two
studies did not find a significant advantage of adjunctive
treatment with HupA. When pooling the three studies
in a random effects meta-analysis, there was no
statistically significant difference in the improvement
in depressive symptoms between MDD patients who
only received antidepressants and those who received
antidepressants and adjunctive HupA.
3.5 Cognitive results
The other results from the three studies are shown in
Table 3. Only two studies[20,21] assessed the cognitive
effects of the treatment. Both studies reported a
significant advantage of using adjunctive HupA. In one
study,[21] memory functioning at the end of the 8-week
trial was better in patients taking antidepressants
with adjunctive HupA than in those who were only
taking antidepressants. In another study,[20] several
measures of executive functioning derived from the
WCST were significantly better at the end of the 6-week
trial in depressed patients taking antidepressants with
adjunctive HupA. These cognitive outcome measures
were quite different so it was not possible to pool the
results of the two studies into a meta-analysis.
3.6 Other results
The level of anxiety was only assessed in one of the
studies.[22] Based on the total score of the HAMA at
the end of the 6-week trial, there was no significant
difference in the severity of anxiety symptoms between
the two groups (Table 3).
10
5
0
5
favors control
Only one study [20] assessed quality of life. As
measured by WHOQOL-100, [15] quality of life was
significantly better at the end of the trial in individuals
who received combined treatment with antidepressants
and HupA (Table 3).
Only one study [22] assessed adverse reactions. The
study assessed adverse events using the DOTES[16] which
considers tachycardia, dysuria, electrocardiographic
abnormality, dry mouth, drowsiness, nausea,
constipation, blurred vision, and insomnia. It found no
difference in the prevalence of adverse events between
the two treatment groups
None of the included RCTs reported the rate or
causes of treatment discontinuation.
4. Discussion
4.1 Main finding
Despite an extensive review of both English-language
and Chinese-language literature, we only identified three
RCTs that assessed the potential benefit of adjunctive
HupA when treating individuals with depression who
are currently using antidepressants. All three studies
were open label and the outcome evaluation in the
trials was not blinded, so the overall strength of the
studies was rated as ‘poor’. The pooled sample from the
three studies, all of which were published in Chinese,
was 238 individuals, but it was only possible to conduct
a meta-analysis for the results related to changes in
depressive symptoms because other outcomes of
interest (e.g., cognitive changes, quality of life changes,
etc.) were only considered in one or two of the studies.
Overall, the results suggest that adjunctive treatment
with HupA over 6 to 8 weeks in patients with depression
who are currently taking antidepressants does not
• 69 •
Table 3. Comparison of cognitive function, anxiety, and quality of life in patients with depression at end of
course of treatment with either antidepressants and adjunctive HupA (experimental group) or with
antidepressants alone (control group)
measure
study
control group
experimental group
n
mean (sd)
n
mean (sd)
t-test (p)
Cognitive measures
WMS-RC
Yang[21]
39
92.1 (16.7)
39
103.0 (15.0)
3.04 (0.003)
WCST (non-perseverative errors)
[20]
Gao
50
35.7 (5.4)
50
27.5 (8.5)
5.71 (<0.001)
WCST (perseverative errors)
Gao[20]
50
37.7 (7.4)
50
26.4 (9.7)
6.60 (<0.001)
WCST (correct responses)
Gao
[20]
50
24.3 (6.2)
50
31.9 (11.3)
4.17 (<0.001)
WCST (categories completed)
Gao[20]
50
3.96 (0.83)
50
4.52 (1.07)
2.92 (0.004)
30
8.1 (7.3)
30
8.3 (7.3)
0.11 (0.909)
50
12.9 (3.9)
50
18.6 (12.5)
3.08 (0.003)
[22]
Anxiety (HAMA total score)
Liu
WHOQOL-100 total score
Gao[20]
[12]
WMS-RC, Wechsler Memory Scale-Revised, Chinese version
WCST, Wisconsin Card Sorting Test[11]
HAMA, Hamilton Anxiety Rating Scale[14]
WHOQOL-100, General Quality of Life Inventory of World Health Organization[15]
result in a better reduction of depressive symptoms,
but it does appear to lead to less cognitive impairment
in depressed individuals and, possibly, to a better selfreported quality of life for depressed individuals.
4.2 Limitations
The small number of studies identified and the limited
measures employed in the identified studies made it
impossible to conduct a full meta-analysis, so we could
not do a sensitivity analysis or subgroup analyses, and
we could not construct a funnel plot to assess potential
publication bias. Specifically, there were not enough
studies with data on cognitive functioning to conduct a
meta-analysis of this important outcome. Moreover, the
relatively low quality of the available studies (open label,
non-blinded) and the relatively short duration of the
studies (from 6 to 8 weeks) means that the findings that
were significant – the benefit of HupA augmentation
for cognitive functioning and quality of life in depressed
patients – are not robust; they need to be replicated in
larger, methodologically more rigorous RCTs that follow
participants for much longer.
4.3 Importance
Despite the limited number of RCTs identified and
the methodological limitations of the identified
studies,[25] this review does provide some support for
the suggestion that AChE inhibitors such as HupA can
ameliorate the cognitive decline that is often associated
with depression and, possibly, improve the quality of
life of individuals being treated for depression with
antidepressant medications. Similar to our findings, a
recent meta-analyses[26] found that adjunctive HupA
is an effective choice for improving cognitive function
in individuals with schizophrenia. The mechanism of
action of HupA in improving cognitive functioning (or
preventing cognitive decline) remains unknown, but
given the importance of cognitive impairment in a wide
range of mental disorders, further work in this promising
area is merited.
Funding
The study was supported by the Start-up Research Grant
(SRG2014-00019-FHS) and the Multi-Year Research
Grant (MYRG2015-00230-FHS) from the University of
Macau. Trial registration number: CRD42015024796
(http://www.crd.york.ac.uk/prospero/)
Conflict of interest statement
The authors report no conflict of interest in conducting
this study and preparing the manuscript.
Authors’ contribution
WZ designed the study and was assisted by YQX and YTX
in the search for papers, data extraction, and analysis.
WZ and YTZ drafted the manuscript. GSU, HFKC, CHN,
and YW made critical revisions to the manuscript. All
authors approved the final version for publication.
• 70 •
石杉碱甲对重度抑郁症患者认知功能障碍的治疗：一项随机对照试验的系统综述
郑伟 , 向应强 , Ungvari GS, Chiu F.K. H, Ng H. C, 王颖 , 项玉涛
背景：乙酰胆碱酯酶 (Acetylcholinesterase, AChE) 抑制
剂在重性抑郁障碍 (Major Depressive Disorder, MDD) 的
动物模型和人类患者中已被证实可以有效地治疗认知
障碍。石杉碱甲 (Huperzine A, HupA) 是一种来自于被称
为蛇足石杉 (Huperzineserrata) 的石松属传统中医药，
是一种强有力的 AChE 抑制剂，已被用于抑郁症的辅助
治疗，但有尚无关石杉碱甲对 MDD 的强化治疗作用的
meta 分析。
目标：对有关石杉碱甲强化治疗抑郁症的随机对照试
验进行系统综述和 meta 分析，评估其疗效及安全性。
方法：两位评估者独立检索 9 个英文和中文数据库，
选择符合预先确定的纳入标准的相关研究，提取有关
疗效和安全性的数据，并进行质量评估和数据拟合合
成。
结果：纳入了三项中国低质量的随机对照试验（总共
n=238），这些试验比较了单用抗抑郁药治疗抑郁症与
抗抑郁药和石杉碱甲的联合治疗，试验中的被试从 16
岁到 60 岁。研究中石杉碱甲辅助抗抑郁药治疗的平均
时间仅为 6.7 周。这三项研究都是公开标签未使用盲
法，所以他们的总体质量评定为差。总体样本的 Meta
分析发现两组抑郁症状的改善没有显著性差异（差
异加权差为 -1.90，95%CI 可信区间为 -4.23 至 0.44，
p=0.11）。然而，石杉碱甲辅助治疗组比单用抗抑郁
药治疗组在认知功能和生活质量方面有显著改善（如
威斯康星卡片分类测验、韦氏记忆量表修订的评估）。
组间药物不良反应的发生率无显著性差异。
结论：有关在接受抗抑郁药的 MDD 患者使用 HupA 辅
助治疗的疗效和安全性的可获取数据不足，难以得出
有关其疗效和安全性的明确结论。汇集国内 3 项低质
量的 RCT 数据没有发现采用辅助使用 HupA 治疗抑郁
症状的优势，但辅助使用 HupA 与更快改善经常伴随
MDD 出现的认知症状相关。
试验注册号码：CRD42015024796 (http://www.crd.york.
ac.uk/prospero/)
关键词：抑郁症；meta 分析；认知功能；石杉碱甲
本文全文中文版从 2016 年 8 月 25 日起在
http://dx.doi.org/10.11919/j.issn.1002-0829.216003 可供免费阅览下载
References
1.
2.
Bhagya V, Srikumar BN, Raju TR, Shankaranarayana Rao BS.
The selective noradrenergic reuptake inhibitor reboxetine
restores spatial learning deficits, biochemical changes,
and hippocampal synaptic plasticity in an animal model of
depression. J Neurosci Res. 2015; 93(1): 104-120. doi: http://
dx.doi.org/10.1002/jnr.23473
Srikumar BN, Raju TR, Shankaranarayana Rao BS.
The involvement of cholinergic and noradrenergic
systems in behavioral recovery following oxotremorine
treatment to chronically stressed rats. Neuroscience.
j.neuroscience.2006.08.041
3.
Pelton GH, Andrews H, Roose SP, Marcus SM, D’Antonio
K, Husn H, et al. Donepezil treatment of older adults with
cognitive impairment and depression (DOTCODE study):
clinical rationale and design. Contemp Clin Trials. 2014; 37(2):
200-208. doi: http://dx.doi.org/10.1016/j.cct.2013.11.015
4.
Islam MR, Moriguchi S, Tagashira H, Fukunaga K. Rivastigmine
improves hippocampal neurogenesis and depression-like
behaviors via 5-HT1A receptor stimulation in olfactory
bulbectomized mice. Neuroscience. 2014; 272: 116-130. doi:
http://dx.doi.org/10.1016/j.neuroscience.2014.04.046
5.
Ago Y, Koda K, Takuma K, Matsuda T. Pharmacological
aspects of the acetylcholinesterase inhibitor galantamine. J
Pharmacol Sci. 2011; 116(1): 6-17
6.
McDermott CL, Gray SL. Cholinesterase inhibitor adjunctive
therapy for cognitive impairment and depressive
symptoms in older adults with depression. The Annals of
pharmacotherapy. 2012; 46(4): 599-605. doi: http://dx.doi.
org/10.1345/aph.1Q445
7.
Matsuda T, Ago Y, Takuma K. [Pharmacological profiles of
galantamine: the involvement of muscarinic receptor].
Nihon shinkei seishin yakurigaku zasshi (Japanese Journal of
Psychopharmacology). 2012; 32(1): 1-8. Japanese
8.
Ma X, Tan C, Zhu D, Gang DR, Xiao P. Huperzine A from
Huperzia species—an ethnopharmacolgical review. J
Ethnopharmacol. 2007; 113(1): 15-34. doi: http://dx.doi.
org/10.1016/j.jep.2007.05.030
9.
Zhang HY, Tang XC. Neuroprotective effects of huperzine
A: new therapeutic targets for neurodegenerative disease.
Trends Pharmacol Sci. 2006; 27(12): 619-625. doi: http://
dx.doi.org/10.1016/j.tips.2006.10.004
10. Xing SH, Zhu CX, Zhang R, An L. Huperzine A in the treatment
of Alzheimer’s disease and vascular dementia: a metaanalysis. Evid Based Complement Alternat Med. 2014; 2014:
363985. doi: http://dx.doi.org/10.1155/2014/363985
11. Kongs SK, Thompson LL, Iverson GL, Heaton RK. Wisconsin
Card Sorting Test-64 Card Version (WCST-64). Odessa, FL:
Psychological Assessment Resources; 2000
12. Chelune GJ, Bornstein RA, Prifitera A. The Wechsler Memory
Scale—Revised. Springer: Advances in Psychological
Assessment; 1990. p. 65-99
13. Hamilton M. A rating scale for depression. Journal of
Neurology, Neurosurgery and Psychiatry; 1960. 23: 56-62
14. Shear MK, Vander Bilt J, Rucci P, Endicott J, Lydiard B, Otto
MW, et al. Reliability and validity of a structured interview
guide for the Hamilton Anxiety Rating Scale (SIGH-A).
Depression & Anxiety. 2001; 13(4): 166–178. doi: http://
dx.doi.org/10.1002/da.1033.abs
15. World Health Organization. WHOQOL: Measuring Quality of
Life. Division of Mental Health and Prevention of Substance
Abuse. World Health Organization; 1997
16. Guy, W. ECDEU assessment manual. In: US Department of
Health. Education and Welfare, Alcohol. Drug Abuse and
Mental Health Administration. Rochville, MD: National
Institute of Mental Health; 1976
17. Higgins JP, Thompson SG. Quantifying heterogeneity in a
meta-analysis. Stat Med. 2002; 21(11): 1539-1558. doi:
http://dx.doi.org/10.1002/sim.1186
18. Der Simonian R, Laird N. Meta-analysis in clinical trials.
Control Clin Trials. 1986; 7(3): 177-188. doi: http://dx.doi.
org/10.1016/0197-2456(86)90046-2
19. Higgins JPT, Green S (eds). Cochrane Handbook for
Systematic Reviews of Interventions. UK, Chichester: John
Wiley & Sons; 2008
20. Gao YF, Li J, Meng HQ , Luo QH, Hu H, Du L. [Effects
of huperzine on cognition function and life quality of
patients with depression]. Chongqing Yi Xue. 2007; 36(6):
483-485. Chinese. doi: http://dx.chinadoi.cn/10.3969/
j.issn.1671-8348.2007.06.001
21. Yang ZB, Deng XM, Zhang GX, Yu XR. [The study of huperzine
combined with fluoxetine on cognition function of patients
with depression]. Lin Chuang Jing Shen Yi Xue Za Zhi. 2010;
20(6): 418-419. Chinese
• 71 •
22.
Liu SZ, Wang PJ, Yin A J, Dang XJ, Guang H. [Effects of
huperzine A combined with venlafaxine for patients
with depression]. Zhongguo Shi Yong Yi Yao. 2010; 5(11):
j.issn.1673-7555.2010.11.116
23. Chinese Medical Association. [Chinese Mental Disorders
Classification and Diagnostic Criteria, Third Edition (CCMD3)]. Jinan: Shandong Science and Technology Press; 2001.
Chinese
24. Sterne JA, Sutton AJ, Ioannidis JP, Terrin N, Jones DR, Lau
J, et al. Recommendations for examining and interpreting
funnel ploy asymmetry in meta-analyses of randomized
controlled trials. BMJ. 2011; 343: d4002. doi: http://dx.doi.
org/10.1136/bmj.d4002
25. Guyatt GH, Oxman AD, Vist GE, Kunz R, Falck-Ytter Y, AlonsoCoello P, et al. GRADE: an emerging consensus on rating
quality of evidence and strength of recommendations. BMJ.
bmj.39489.470347.AD
26. Zheng W, Xiang YQ, Li XB, Ungvari GS, Chiu HFK, Sun F,
et al. Adjunctive huperzine A for cognitive deficits in
schizophrenia: a systematic review and meta-analysis. Hum
Psychopharmacol: Clinical and Experimental. 2016; doi:
10.1002/hup.2537
(received, 2016-01-11; accepted, 2016-03-20)
Dr. Wei Zheng obtained a bachelor’s degree from Hebei Medical University in 2012 and a master’s
degree of psychiatry from Capital Medical University in Beijing in 2015. He is currently a resident
psychiatrist in the Department of Psychiatry at the Affiliated Brain Hospital of Guangzhou Medical
University (Guangzhou Huiai Hospital) in Guangdong Province, China.
• 72 •
•Original research article•
A community-based controlled trial of a comprehensive
psychological intervention for community residents with
diabetes or hypertension
Qingzhi ZENG1, Yanling HE1,*, Zhenyu SHI2, Weiqing LIU3, Hua TAO4, Shiming BU5, Donglei MIAO6,
Ping LIU7, Xuanzhao ZHANG8, Xiaoping LI9, Xuejun QI10, Qin ZHOU11
Background: Depression and anxiety often occur in persons with chronic physical illnesses and typically
magnify the impairment caused by these physical conditions, but little attention has been paid to this issue
in low- and middle-income countries.
Aim: Evaluate the effectiveness of a community-based psychological intervention administered by nonspecialized clinicians and volunteers for alleviating depressive and anxiety symptoms in individuals with
chronic physical illnesses.
Methods: A total of 10,164 community residents receiving treatment for diabetes or hypertension in
Shanghai were arbitrarily assigned to a treatment-as-usual condition (n=2042) or an intervention condition
(n=8122) that included community-wide psychological health promotion, peer support groups, and individual
counseling sessions. The self-report Patient Health Questionnaire (PHQ-9), Generalized Anxiety Disorder
scale (GAD-7), and 12-item Short-Form Health Survey (SF-12) assessed depressive symptoms, anxiety
symptoms, and quality of life at baseline and after the 6-month intervention.
Results: Among the 8813 individuals who completed the baseline assessment, 16% had mild or more
severe depressive or anxiety symptoms (PHQ-9 or GAD-7 >5) and 4% had moderate or severe depressive
or anxiety symptoms (PHQ-9 or GAD-7 >10). The education component of the intervention was effectively
implemented, but only 31% of those eligible for peer-support groups and only 9% of those eligible for
individual counseling accepted these interventions. The dropout rate was high (51%), and there were
significant differences between those who did and did not complete the follow-up assessment. After
adjusting for these confounding factors, the results in individuals who completed both assessments indicated
that the intervention was associated with significant improvements in depressive symptoms (F=9.98,
p<0.001), anxiety symptoms (F=12.85, p<0.001), and in the Mental Component Summary score of the SF-12
(F=16.13, p<0.001). There was, however, no significant change in the self-reported rates of uncontrolled
diabetes or hypertension.
Conclusions: These results support the feasibility of implementing community-based interventions to
reduce the severity of depressive and anxiety symptoms in persons with chronic medical conditions in lowand middle-income countries where psychiatric manpower is very limited. However, there are substantial
methodological challenges to mounting such interventions that need to be resolved in future studies before
the widespread up-scaling of this approach will be justified.
Keywords: depression; anxiety; community intervention; diabetes; hypertension; community medical
service; China
1
Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai, China
Shanghai Pudong New Area Mental Health Center, Shanghai, China
3
Xinhua Community Health Center of the Changning District, Shanghai, China
4
Changning District Mental Health Center, Shanghai, China
5
Minhang District Mental Health Center, Shanghai, China
6
Jiangsu Community Health Center of the Changning District, Shanghai, China
7
Xinzhuang Community Health Center of the Minhang District, Shanghai, China
8
Jiangchuan Community Health Center of the Minhang District, Shanghai, China
9
Corning Hospital, Shenzhen, China
10
Hangzhou Seventh People’s Hospital, Hangzhou, China
11
Fudan University School of Public Health, Shanghai, China
2
*correspondence: Professor Yanling He, Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, 600 Wan Ping Nan Road,
Shanghai 200030, China. E-mail: [email protected]
1. Introduction
Diabetes and hypertension are two common chronic
illnesses that are quite prevalent in China: 26.7% of the
adult population (265 million individuals) has primary
hypertension[1] and 11.6% (110 million individuals) has
adult-onset diabetes.[2] Studies in other countries report
that individuals with diabetes and hypertension are
more likely to have depressive disorders and anxiety
disorders than individuals without these physical
illnesses. [3,4] Moreover, compared to persons with
hypertension or diabetes who do not have comorbid
depression or anxiety, those with comorbid depression
or anxiety are less likely to adhere to medication
regimens, have a lower quality of life, experience an
earlier onset of complications, and have higher mortality
rates and higher medical costs.[5,6]
Several studies report the effectiveness of
psychological interventions for depression and anxiety
in individuals with diabetes or hypertension. [7,8]
However, most of these studies suffer from significant
limitations: they (a) are targeted to the relatively small
number of individuals who meet diagnostic criteria of
major depressive disorder or anxiety disorder, excluding
the much larger number of individuals with mild to
moderate depressive and anxiety symptoms; (b) involve
a single type of individual-based treatment (medication,
cognitive behavioral therapy, etc.) that requires a
high level of expertise to administer; (c) focus on the
reduction of depressive or anxiety symptoms with little
consideration of other important outcomes such as
quality of life, changes in the severity of the physical
disorder, overall treatment costs, and family burden;
and (d) have sample sizes that are too small and too
unrepresentative to assess the effect of the intervention
on all community members with hypertension or
diabetes.
In China little attention has been paid to comorbid
depressive and anxiety symptoms in persons with
hypertension or diabetes, but the impression for the
limited research on the issue is that sub-threshold forms
of depression or anxiety (i.e., episodes that do not
meet full diagnostic criteria) are much more common
than full-blown episodes of major depressive disorder
or generalized anxiety disorder.[9] Community-based
health services in China do not have the resources or
personnel needed to provide sophisticated, individualbased psychopharmacological or psychotherapeutic
services to these individuals, so we decided to adapt the
multi-faceted ‘Collaborative Care Model,’[10,11] originally
developed in the United States, for use in Shanghai.
This care-delivery model is targeted at all patients with
hypertension or diabetes, regardless of the severity of
their psychological symptoms. It aims to improve service
quality by creating community-based health care teams
that integrate routine surveillance and positive followup of patients’ medical condition with assessment of
their psychological status, and, if necessary, provision of
social support to help the individual and his/her family
members adjust to their stressful life circumstances.
The current study uses a community-based design
to assess the effectiveness of this comprehensive
• 73 •
approach to improve the psychological health, physical
health, and quality of life of individuals with diabetes or
hypertension.
2. Methods
Community health services in Shanghai are provided
by community health centers (CHCs) distributed
throughout the municipality’s 16 districts. Each
community health center has a number of ‘community
health service teams’ responsible for monitoring chronic
illnesses among residents of several neighborhoods
within the service area covered by the community
health center. Each service team typically includes a
general doctor, a nurse, and a public health clinician;
among other responsibilities, they are expected to
establish and maintain a registry of all residents with
hypertension or diabetes in the neighborhoods; assess
their blood pressure, blood sugar, and medication
adherence at least four times a year; provide a full
medical exam annually; refer those who need more
advanced treatment; and provide related health
education.
2.1 Sample
Study participants were community residents registered
with diabetes or hypertension from three CHCs in two of
Shanghai’s 16 districts (the Xinhua CHC and the Huayang
CHC in the Changning District and the Xinzhuang CHC in
the Minhang District). As shown in Figure 1, participants
came from 62 neighborhoods in the catchment areas
of these three CHCs that were provided services by
11 separate community health service teams; all 17
neighborhoods serviced by four community health
service teams in the Xinhua CHC; all 21 neighborhoods
serviced by four community health service teams in
the Huayang CHC; and 24 of the 55 neighborhoods
serviced by three of the community health teams in the
Xinzhuang CHC. The study inclusion criteria for residents
of these communities were as follows: (a) aged 18 years
or older; (b) resided in the community; (c) registered at
the community health center with a diagnosis of adultonset diabetes or primary hypertension (typically these
conditions are initially diagnosed at a general hospital
outpatient department and then referred back to the
CHC for follow-up care); (d) no physical illness that was
so severe it made it impossible to participate; (e) no
mental disorder or cognitive impairment that made it
impossible to participate; and (f) provided written or
oral informed consent to participate in the study.
We estimated the sample size based on the
prevalence of clinically significant depressive and
anxiety symptoms. Assuming a relatively conservative
mean baseline prevalence of 15%, in order to observe a
20% improvement (mean prevalence drop to 12%), a 3:1
ratio of intervention and control subjects, a type I error
rate of 5% (i.e., α<0.05), a type II error rate of 80% (i.e.,
β>0.80), and a 30% dropout rate over the 6 months of
follow-up, there needed to be at least 4233 participants
in the intervention group and 1409 participants in the
control group.
• 74 •
Figure 1. Flowchart of the study
Community residents from 62 neighborhoods in two of Shanghai’s 16 districts (Changning District and Minhang
District) were provided follow-up management of chronic illnesses by 11 community health service teams from
three community health centers (CHCs), (Xinhua CHC, Huayang CHC, and Xinzhuang CHC), from August 2012 to
December 2013
Health service teams (and the neighborhoods managed by each team) were arbitrarily assigned to the
intervention group or control group based on the estimated number of participants needed in each group
13,338 residents from 34 neighborhoods were
provided health services by 6 service teams
working out of the Xinhua CHC and the
Xinzhuang CHC
10,244 residents from 28 neighborhoods were provided
health services by 5 service teams working out
of the Huayang CHC and the Xinzhuang CHC
8122 individuals with diabetes or hypertension were
assigned to the intervention groupa
2042 individuals with diabetes or hypertension were
assigned to the control groupb
6897 completed the baseline measure:
• 6897 completed PHQ-9 and GAD-7
• 6866 completed SF-12
1916 complete the baseline measure:
• 1916 completed PHQ-9, GAD-7, and SF-12
6897 received routine community management of
chronic illness plus a 6-month comprehensive
psychological intervention:
• 6897 received mass health education
• 325 attended peer support groups
•
24 attended individual sessions of Problem
Solving Treatment for Primary Care
2042 received routine community management of
chronic illness
5561 individuals with diabetes or hypertension from
19 of the 34 neighborhoods were selected for
outcome assessment of the intervention groupa
3694 individuals in the intervention group completed
the outcome evaluation:
• 3694 completed PHQ-9 (100%)
• 3694 completed GAD-7 (100%)
• 3577 completed SF-12 (97%)
• 3015 completed blood pressure measure (82%)
• 2979 completed blood sugar measure (81%)
1394 individuals in control group completed outcome
evaluation
• 1394 completed PHQ-9 (100%)
• 1394 completed GAD-7 (100%)
• 1353 completed SF-12 (97%)
• 1225 completed blood pressure measure (88%)
• 1210 completed blood sugar measure (87%)
3039 intervention group subjects completed both
evaluations
1239 control group subjects completed both
evaluations
PHQ-9, Patient Health Questionnaire-9[13]
GAD-7, 7-item Generalized Anxiety Disorder Scale[14]
SF-12, 12-Item Short-Form Health Survey[15]
a
individuals with diabetes were all included while those with hypertension were randomly selected
b
proportion of diabetes and hypertension patients selected to match the proportion in the baseline
intervention group
Based on the number of registered individuals with
diabetes and hypertension in the neighborhoods in
the catchment areas of the three participating CHCs,
we arbitrarily assigned the 11 community health
service teams from the CHCs to the intervention group
or the control group such that the ratio of potential
subjects in the intervention and control groups was
approximately 3 to 1. As shown in Figure 1, the active
psychological intervention and standard follow-up care
(the intervention group) were provided to residents
of 34 neighborhoods (17 neighborhoods provided
services by four service teams from Xinhua CHC and 17
neighborhoods provided services by two service teams
from Xinzhuang CHC) and standard follow-up care
alone (the control condition) was provided to residents
of 28 neighborhoods (21 neighborhoods provided
services by four service teams from Huayang CHC and 7
neighborhoods provided services by one service team
from Xinzhuang CHC).
Research studies indicate that the relationship
between diabetes and depressive or anxiety symptoms
is stronger than that between hypertension and
depressive or anxiety symptoms, [12] so we included
all individuals with diabetes from the intervention
communities and then increased the sample to the
desired size by taking a simple random sample from
the residents with hypertension. Based on the ratio of
diabetes and hypertension among individuals eligible
for the intervention group, corresponding proportions
of diabetes and hypertension patients were randomly
selected from all diabetes and hypertension patients
living in the control communities. After the 6-month
intervention, limited resources and personnel made
it impossible to redo the evaluation of all intervention
group participants, so 19 of the 34 neighborhoods in the
intervention group were selected (those that were most
active in implementing the psychological intervention),
and all persons registered with diabetes or hypertension
from these neighborhoods were selected for follow-up
evaluation. In the control neighborhoods, all individuals
assessed at baseline were selected for the 6-month
follow-up evaluation.
2.2 Intervention
All participants received routine management of
their chronic illness. As described above, in CHCs
in Shanghai this is officially supposed to include
registration, complete annual physical examinations,
and quarterly follow-up of community residents
with adult-onset diabetes and primary hypertension.
The quarterly follow-up assessments include
assessment of blood pressure and fasting blood
glucose, identification of sequelae or comorbid health
conditions, health education about lifestyle issues,
medication management, and, if necessary, referral
to hospital outpatient or inpatient services for more
extensive evaluation or treatment. The degree to which
community residents with diabetes and hypertension
participate in these CHC services varies considerably.
The community-based comprehensive psychological
intervention used in this study was an adaptation of
• 75 •
the IMPACT model developed in the United States
for use in Shanghai. [10,11] In addition to the routine
management of their diabetes and/or hypertension, all
intervention group subjects also received communitybased education about psychological health. Some
individuals in the intervention group also received
additional psychological support: individual counseling
was offered to individuals whose baseline scores on
the Patient Health Questionnaire-9 (PHQ-9)[13] (which
evaluates depressive symptoms) or the Generalized
Anxiety Disorder 7-item scale (GAD-7)[14] were >10; and
small-group peer support was offered to individuals
whose total score on either scale was >5.
The community-based mental health education
co m p o n e nt i nvo l ve d d i st r i b u t i n g b ro c h u re s ,
broadcasting educational videos, and hosting lectures
about psychosomatic health for individuals with chronic
illnesses. The content focused on the identification
and management of the symptoms of depression and
anxiety, the relationship between psychological health
and somatic health, and the relationship between stress
and depression or anxiety.
The peer support group intervention targeted
patients with diabetes or hypertension who had PHQ-9
or GAD-7 scores > 5 but also welcomed the participation
of other community members who expressed interest
in the groups. This intervention involved monthly 6090 minute meetings led by community volunteers who
had received guidance from counselors. The group
meetings, which typically included 9-18 individuals,
focused on (a) the management of chronic diseases,
(b) healthy lifestyles, (c) psychological coping skills for
dealing with diabetes and hypertension, (d) knowledge
about depression and anxiety, and (e) self-awareness
of negative emotions. In addition to the transmission
of crucial information, the meetings also provided
emotional and social support to the participants,
something that previous research has shown to reduce
depressive symptoms and improve the control of
diabetes and hypertension.[15]
The individual intervention targeted individuals
whose PHQ-9 or GAD-7 score was >10. Counselors
(individuals who had a nationally approved Level-2
counseling certificate) provided one 60-minute and
six 30-minute sessions of Problem Solving Treatment
for Primary Care (PST-PC)[16] to each individual. The
counseling focused on alleviating symptoms of
depression and anxiety by assisting these individuals to
become more self-aware, to learn how to analyze and
deal with their problems, to decrease their feelings of
frustration, and to increase their feelings of control over
their lives. PST has been found to be effective in the
management of emotional problems among patients
treated at community health centers.[16]
The three components of this communitybased intervention in the 34 neighborhoods was
collaboratively coordinated and provided by 391
individuals, including local administrators, community
clinicians, community public health workers, counselors,
and volunteers. All individuals who provided each of
the three components of the intervention received
• 76 •
appropriate training before implementing the
intervention. We ensured that the group leaders
and counselors grasped related skills through the
introduction of learning theories, the illustration of
examples, discussion, and role-play exercises. During
the intervention process, peer support leaders and
the counselors also routinely received professional
supervision in order to identify and address any
problems in a timely manner.
2.3 Measures
At baseline all participants completed a detailed
demographic and clinical status form, the PHQ-9[13]
to assess the severity of depression, the GAD-7[14] to
assess the severity of anxiety, and the 12-Item ShortForm Health Survey (SF-12)[17] to assess quality of life.
Six months later the PHQ-9, GAD-7, and SF-12 were readministered, and participants were asked to classify
the control of their diabetes and/or hypertension as
‘very stable’, ‘stable’, or ‘unstable’.
Demographic and clinical variables considered
included age, gender, marital status, level of education,
employment status, age of onset of current illness,
course of illness, presence of physical sequelae of
diabetes of hypertension, and frequency of hospitalbased treatment (as outpatient or inpatient) in the prior
6 months.
The PHQ-9 and GAD-7 are widely used selfcompletion scales with good reliability and validity[18,19]
which assess the frequency of specific depressive
and anxiety symptoms over the prior two weeks.
The items on both scales are rated on 4-point Likert
scales (0=’never’ to 3=’almost every day’), so the
total score for 9-item PHQ-9 ranges from 0 to 27
and that for 7-item GAD-7 ranges from 0 to 21, with
higher scores representing more severe depressive or
anxiety symptoms. The PHQ-9 total score is classified
as follows: [18] 0 to 4, ‘no depression’; 5 to 9, ‘mild
depression’; 10 to 14, ‘moderate depression’; 15 to 19,
‘moderate to severe depression’; 20 or above, ‘severe
depression’. The GAD-7 total score is classified as
follows:[19] 0 to 4, ‘no anxiety’; 5 to 9, ‘mild anxiety’; 10
to 14, ‘moderate anxiety’; 15 or above, ‘severe anxiety’.
Research has shown that the SF-12[17] is a valid
measure of quality of life in the general Chinese
population.[20] We use two components from the scale in
the current analysis: the Mental Component Summary
(MCS) score and the Physical Component Summary (PCS)
score. These scores are based on weighting responses to
all 12 items, with higher scores indicating better quality
of life.
2.5 Statistical analysis
We used EpiData 3.1 (The EpiData Association, Odense,
Denmark) to input and manage the data and used SPSS
17.0 (SPSS Inc., Chicago, IL, USA) to analyze the data.
Categorical data were compared using Chi-square tests,
continuous data were analyzed using parametric or nonparametric tests depending on whether or not the data
was distributed normally.
The main analysis was based on the subset of
participants who completed both the baseline and
6-month evaluations. Six subgroups of respondents
were identified according to the baseline results on the
PHQ-9 and GAD-7: (1) those with PHQ-9 >5; (2) those
with GAD-7 >5; (3) those with PHQ-9 >10; (4) those with
GAD-7 >10; (5) those with PHQ-9 or GAD-7 >5; and (6)
those with PHQ-9 or GAD-7 >10.
3. Results
3.1 Completion status
There were 10,164 individuals with diabetes or
hypertension registered in the 62 participating
communities and 8813 of them (86.7%) completed the
baseline evaluation; 6897 of the 8122 (84.9%) residents
in the intervention group neighborhoods with diabetes
or hypertension completed the baseline assessment
and 1916 of the 2042 (93.9%) residents in the control
group neighborhoods with diabetes or hypertension
completed the baseline assessment. The main reasons
for failure to participate in the study were failure to
meet the inclusion criteria, refusal to participate, and
difficulty of access to the CHC (some registered residents
at the CHCs actually live elsewhere). Comparison of the
1351 who did not participate with the 8813 who did
participate found no significant difference by gender
(46.7% v. 45.2% male, respectively, X2=1.02, p=0.314)
or in the mean (sd) age (70.0 [10.2] v. 69.6 [10.3] years,
respectively, t=1.14, p=0.253).
Only 19 of the 34 intervention communities
participated in the 6-month outcome evaluation, but
all 28 control communities participated in the 6-month
follow-up evaluation. In total 7603 individuals were
selected to participate in the outcome evaluation and
5088 of them (66.9%) completed the evaluation; in the
intervention group 3694 of the 5561 (66.4%) selected
individuals completed the outcome assessment and in
the control group 1394 of the 2042 (68.3%) selected
individuals completed the outcome assessment.
As shown in Figure 1, 3039 participants in the
intervention group and 1239 in the control group
completed both the baseline and the outcome
evaluations.
3.2 Comparison of individuals who do and do not
complete both evaluations
Table 1 compares the demographic and clinical
characteristics of individuals in the control group
and the intervention group who only completed
the baseline evaluation with the characteristics of
individuals from the two groups who completed both
the baseline and 6-month follow-up evaluations (and
thus, were included in the outcome assessment for
the intervention). In the control group, the mean
(sd) age of the 1239 individuals who completed
both evaluations was not significantly different from
that of the 677 individuals who only completed the
baseline assessment (70.4 [10.3] v. 69.6 [10.1] years,
respectively, t=1.08, p=0.279), but individuals who
• 77 •
Table 1. Comparison of demographic characteristics and illness characteristics in the intervention group and
the control group between respondents who only completed the baseline assessment and those
who completed both the baseline and the 6-month outcome assessmenta
control group
only
only
completed completed
completed completed
both
baseline
baseline
both
X2
X2
assessments
assessment
assessment assessments
(n=677)
(n=1239) (p-value)
(n=3858)
(n=3039) (p-value)
n (%)
n (%)
n (%)
n (%)
characteristic
age group
<65 years
65- 80
>80
male
female
retired
employment
working/studying
status
other
institution manager
professional/technician
general worker
occupation
laborer
other
illiterate
elementary school
educational
level
middle school
college degree
never married
marital
married
status
divorced/widowed
only hypertension
illness
only diabetes
hypertension + diabetes
0 no sequelae
sequelae of
diabetes or
1 sequela
hypertension
2+ sequelae
<6 years
years duration of 6-10 years
illness
>11 years
0 hospital visits
hospital-based
treatments in the 1-2 hospital visits
last 6 months
3+ hospital visits
gender
a
intervention group
215 (31.8)
343 (50.7)
119 (17.6)
374 (30.2)
623 (50.3)
242 (19.5)
1.25
(0.535)
1302 (33.7) 1020 (33.6)
1851 (48.0) 1464 (48.2)
705 (183.) 555 (18.3)
0.03
(0.985)
302 (44.6)
375 (55.4)
611 (90.4)
35 (5.2)
30 (4.4)
117 (17.3)
170 (25.1)
105 (15.5)
263 (38.9)
21 (3.1)
66 (9.8)
111 (16.4)
376 (55.6)
123 (18.2)
9 (1.3)
539 (79.9)
127 (18.8)
396 (58.5)
100 (14.8)
181 (26.7)
419 (61.9)
166 (24.5)
92 (13.6)
159 (23.5)
172 (25.4)
345 (51.0)
466 (69.2)
49 (7.3)
158 (23.5)
564 (45.5)
675 (54.5)
1138 (91.9)
69 (5.6)
31 (2.5)
199 (16.1)
258 (20.8)
223 (18.0)
532 (42.9)
27 (2.2)
77 (6.2)
213 (17.2)
723 (58.4)
225 (18.2)
17 (1.4)
1023 (82.6)
198 (16.0)
674 (54.4)
192 (15.5)
373 (30.1)
783 (63.2)
277 (22.4)
178 (14.4)
285 (23.0)
322 (26.0)
631 (51.0)
860 (69.6)
131 (10.6)
245 (19.8)
0.15
(0.701)
5.37
(0.068)
1770 (45.9)
2088 (54.1)
3488 (90.4)
249 (6.5)
120 (3.1)
605 (15.7)
951 (24.7)
757 (19.6)
1415 (36.7)
125 (3.2)
217 (5.6)
595 (15.4)
2292 (59.4)
752 (19.5)
45 (1.2)
3280 (85.1)
531 (13.8)
2341 (60.7)
449 (11.6)
1068 (27.7)
2590 (67.2)
771 (20.0)
493 (12.8)
887 (24.2)
1024 (27.9)
1758 (47.9)
2940 (76.6)
320 (8.3)
580 (15.1)
1.39
(0.238)
0.002
(0.999)
8.84
(0.065)
8.10
(0.044)
2.47
(0.291)
3.18
(0.203)
1.19
(0.551)
0.102
(0.950)
7.85
(0.020)
1351 (44.5)
1688 (55.5)
2748 (90.5)
196 (6.5)
94 (3.1)
536 (17.6)
824 (27.1)
502 (16.5)
1093 (36.0)
83 (2.7)
152 (5.0)
455 (15.0)
1732 (57.0)
698 (23.0)
22 (0.7)
2544 (83.7)
473 (15.6)
2122 (69.8)
224 (7.4)
693 (22.8)
2143 (70.5)
573 (18.9)
322 (10.6)
746 (25.3)
712 (24.2)
1487 (50.5)
2322 (76.9)
205 (6.8)
492(16.3)
18.60
(0.001)
12.93
(0.005)
7.56
(0.023)
69.56
(<0.001)
10.80
(0.005)
11.77
(0.003)
6.82
(0.033)
MISSING DATA FOR RESPONDENTS IN THE CONTROL GROUP: for those who only completed the baseline assessment, there were 1
missing data in employment status, 1 in occupation, 1 in educational level, 2 in marital status, 1 in years duration of illness, and 4 in
hospital-based treatments in the last 6 months; for those who completed both assessments, there were 1 missing data in employment
status, 1 in educational level, 1 in marital status, 1 in sequelae of diabetes or hypertension, 1 in years duration of illness, and 3 in
hospital-based treatments in the last 6 months;
MISSING DATA FOR RESPONDENTS IN THE INTERVENTION GROUP: for those who only completed the baseline assessment, there were
1 missing data in employment status, 5 in occupation, 2 in education level, 2 in marital status, 4 insequelae of diabetes or hypertension,
189 in years duration of illness, and 18 in hospital-based treatments in the last 6 months; for those who completed both assessments,
there were 1 missing data in employment status, 1 in occupation, 2 in education level, 1 in sequelae of diabetes or hypertension, 94 in
years duration of illness, and 20 in hospital-based treatments in the last 6 months
• 78 •
completed both assessments had a higher level of
education and had made fewer hospital visits for
treatment of their diabetes and/or hypertension in the
prior 6 months than individuals who only completed
the baseline assessment. In the intervention group,
there was also no significant difference in age between
the 3039 individuals who completed both assessments
compared to that of the 3858 individuals who only
completed the baseline assessment (69.4 [10.3] v. 69.4
[10.3] years, respectively, t=0.11, p=0.916), but several
other variables were significantly different between the
two subgroups of individuals living in the intervention
group neighborhoods: compared to individuals who
only completed the baseline assessment, those who
completed both assessments were more likely to
be professionals or managers, had a higher level of
education, were more likely to be divorced or widowed,
were more likely to only have hypertension, were less
likely to have complications (sequelae) of diabetes or
hypertension, had a longer duration of illness, and were
more likely to have made multiple hospital visits for the
management of their illness over the prior 6 months.
Comparison of the baseline results for the four
primary outcome measures between those who
only completed the baseline evaluation and those
who completed both evaluations was as follows. In
the control group the mean (sd) PHQ-9 for the 1239
individuals who completed both evaluations and the 677
individuals who only completed the baseline evaluation
were 2.39 (3.42) and 2.26 (3.60), respectively (t=-0.82,
p=0.414); the corresponding results for the GAD-7
were 1.16 (2.36) and 1.12 (2.59) (t=-0.37, p=0.710);
those for the PCS of the SF-12 were 45.0 (8.9) and 45.1
(9.5), (t=0.30, p=0.765); and those for the MCS of the
SF-12 were 54.4 (8.8) and 55.2 (9.1) (t=1.75, p=0.081).
In the intervention group the mean (sd) PHQ-9 for the
3039 individuals who completed both evaluations and
the 3858 individuals who only completed the baseline
evaluation were 1.90 (3.17) and 2.18 (3.45), respectively
(t=3.46, p=0.001); the corresponding results for the
GAD-7 were 0.88 (2.11) and 1.10 (2.54) (t=3.89,
p<0.001); those for the PCS of the SF-12 were 46.2 (8.4)
and 45.5 (9.0) (t=-3.52, p<0.001); and those for the MCS
of the SF-12 were 55.5 (8.3) and 54.1 (8.4), respectively
(t=-7.02, p<0.001).
3.3 Comparison of characteristics of the two groups at
baseline and after both assessments
Table 2 shows the comparison of the baseline
demographic and clinical variables for individuals who
completed the baseline evaluation in the intervention
and control groups and for individuals who completed
both the baseline and 6-month follow-up evaluations in
the two groups. At baseline, there were no significant
differences between the intervention and control groups
by gender, employment status, or duration of illness,
but, given the very large sample, several relatively small
differences between the groups in other variables were
statistically significant. For example, the mean (sd)
age in the control group was 70.5 (10.2) years versus
69.7 (10.3) years in the intervention group; this minor
difference in mean age of 0.8 years was statistically
significant (t=9.18, p=0.002). As shown in the table,
compared to control group participants, intervention
group participants were also significantly less likely to
be manual laborers (36.4% v. 41.5%), more likely to have
a college education (21.0% v. 18.2%), more likely to be
married (84.5% v. 81.7%), much more likely to only have
hypertension (64.7% v. 55.8%), less likely to have one
or more sequelae of diabetes or hypertension (31.3%
v. 37.2%), and less likely to have made one or more
hospital visits (as outpatient or inpatient) to manage
their illness in the prior 6 months (23.3% v. 30.5%).
Most of the differences between the intervention
and control groups seen at the baseline assessment
persisted in the subgroup of individuals who completed
both baseline and follow-up assessments. Compared
to control group participants, intervention group
participants were less likely to be manual laborers, more
likely to have a college education, much more likely to
only have hypertension, less likely to have one or more
sequelae of diabetes or hypertension, and less likely
to have made one or more hospital visits to manage
their illness in the prior 6 months. Intervention group
participants who completed both evaluations were also
younger than control group participants who completed
both evaluations (69.4 [10.2] v. 70.4 [10.3] years,
respectively, t=2.97, p=0.003).
3.4 Prevalence of depressive and anxiety symptoms at
baseline
Combining the results of all 8813 community residents
with hypertension or diabetes who completed the
baseline assessment with PHQ-9 and the GAD-7
from both the intervention and control groups, the
prevalence of the six categories of depressive and
anxiety conditions were as follows: 14.7% (1292/8813)
had mild or more severe depressive symptoms
(PHQ-9 >5); 7.0% (613/8813) had mild or more severe
anxiety symptoms (GAD-7 >5); 16.0% (1409/8813) had
mild or more severe depressive or anxiety symptoms
(PHQ-9 or GAD-7 >5); 3.9% (344/8813) had moderate
or severe depressive symptoms (PHQ-9 >10); 1.6%
(140/8813) had moderate or severe anxiety symptoms
(GAD-7 >10); and 4.2% (369/8813) had moderate or
severe depressive or anxiety symptoms (PHQ-9 or
GAD-7 >10).
The 8813 individuals who completed the baseline
assessments included 5533 with primary hypertension
only, 965 with adult-onset diabetes only, and 2315 with
both hypertension and diabetes. The prevalence of mild
or more severe depressive or anxiety symptoms (PHQ-9
or GAD-7 >5) in these three groups of respondents
was 13.4%, 17.7%, and 21.3%, respectively (X2=78.11,
df=2, p<0.001). The prevalence of moderate or severe
depressive or anxiety symptoms (PHQ-9 or GAD-7 >10)
in the three groups of respondents was 3.3%, 4.9%, and
6.0%, respectively (X2=29.52, df=2, p<0.001).
• 79 •
Table 2. Comparison of demographic characteristics and illness characteristics between the intervention
group and the control group at baseline and among individuals who completed the baseline and the
6-month assessmentsa
characteristic
age group
gender
employment
status
<65 years
65- 80
>80
male
female
retired
working/studying
other
institution manager
professional/technician
general worker
occupation
laborer
other
illiterate
elementary school
educational
middle school
level
college degree
never married
marital status married
divorced/widowed
only hypertension
completed baseline assessment
control intervention
group
group
X2
(n=1916)
(n=6897)
(p-value)
n (%)
n (%)
589 (30.7) 2322 (33.7)
5.86
966 (50.4) 3315 (48.1)
(0.053)
361 (18.8) 1260 (18.3)
866 (45.2) 3121 (45.3)
0.002
1050 (54.8) 3776 (54.7)
(0.967)
1749 (91.4) 6236 (90.4)
2.68
104 (5.4)
445 (6.5)
(0.262)
61 (3.2)
214 (3.1)
316 (16.5)
428 (22.3)
328 (17.1)
795 (41.5)
48 (2.5)
143 (7.5)
324 (16.9)
1099 (57.4)
348 (18.2)
26 (1.4)
1562 (81.7)
325 (16.9)
1070 (55.8)
292 (15.2)
Illness
only diabetes
hypertension + diabetes 554 (28.9)
1,202 (62.8)
0 no sequelae
sequelae of
443 (23.1)
diabetes or
1 sequela
hypertension 2+ sequelae
270 (14.1)
444 (23.2)
<6 years
years duration 6-10 years
494 (25.8)
of illness
976 (51.0)
>11 years
hospital-based 0 hospital visits
1326 (69.5)
treatments
180 (9.4)
1-2
hospital
visits
in the last 6
403 (21.1)
months
3+ hospital visits
a
1141 (16.6)
1775 (25.8)
1259 (18.3)
2508 (36.4)
208 (3.0)
369 (5.4)
1050 (15.2)
4024 (58.4)
1450 (21.0)
67 (1.0)
5824 (84.5)
1004 (14.5)
4463 (64.7)
673 (9.8)
1761 (25.5)
4,733 (68.7)
1,344 (19.5)
815 (11.8)
1633 (24.7)
1736 (26.2)
3245 (49.1)
19.87
(<0.001)
5262 (76.7)
525 (7.7)
1072 (15.6)
42.99
(<0.001)
20.53
(<0.001)
9.39
(0.009)
66.45
(<0.001)
23.79
(<0.001)
2.58
(0.275)
completed both assessments
control intervention
group
group
X2
(n=1239)
(n=3039)
(p-value)
n (%)
n (%)
374 (30.2) 1020 (33.6)
4.65
623 (50.3) 1464 (48.2) (0.098)
242 (19.5) 555 (18.3)
564 (45.5) 1351 (44.5)
0.40
675 (54.5) 1688 (55.5) (0.525)
1138 (91.9) 2748 (90.5)
2.35
69 (5.6)
196 (6.5)
(0.309)
31 (2.5)
94 (3.1)
199 (16.1)
258 (20.8)
223 (18.0)
532 (42.9)
27 (2.2)
77 (6.2)
213 (17.2)
723 (58.4)
225 (18.2)
17 (1.4)
1023 (82.6)
198 (16.0)
674 (54.4)
192 (15.5)
373 (30.1)
783 (63.2)
277 (22.4)
178 (14.4)
285 (23.0)
322 (26.0)
631 (51.0)
860 (69.6)
131 (10.6)
245 (19.8)
536 (17.6)
824 (27.1)
502 (16.5)
1093 (36.0)
83 (2.7)
152 (5.0)
455 (15.0)
1732 (57.0)
698 (23.0)
22 (0.7)
2544 (83.7)
473 (15.6)
2122 (69.8)
224 (7.4)
693 (22.8)
2143 (70.5)
573 (18.9)
322 (10.6)
746 (25.3)
712 (24.2)
1487 (50.5)
2322 (76.9)
205 (6.8)
492(16.3)
28.48
(<0.001)
14.91
(0.002)
4.29
(0.117)
110.64
(<0.001)
23.04
(<0.001)
3.11
(0.211)
28.71
(<0.001)
MISSING DATA FOR ALL RESPONDENTS WHO COMPLETED THE BASELINE ASSESSMENT; in the control group, there were 2 missing data in employment
status, 1 in occupation, 2 in educational level, 3 in marital status, 1 in sequelae of diabetes or hypertension, 2 in years duration of illness, and 7 in
hospital-based treatments in the last 6 months; and for those in the intervention group there were 2 missing data in employment status, 6 in occupation,
4 in education level, 2 in marital status, 5 in sequelae of diabetes or hypertension, 283 in years duration of illness, and 38 in hospital-based treatments in
the last 6 months
MISSING DATA FOR RESPONDENTS WHO COMPLETED BOTH ASSESSMENTS; in the control group, there were 1 missing data in employment status, 1
in educational level, 1 in marital status, 1 insequelae of diabetes or hypertension, 1 in years duration of illness, and 3 in hospital-based treatments in
the last 6 months; and for those in the intervention group there were 1 missing data in employment status, 1 in occupation, 2 in education level, 1 in
sequelae of diabetes or hypertension, 94 in years duration of illness, and 20 in hospital-based treatments in the last 6 months
• 80 •
3.5 Fidelity of the implementation of the communitybased psychological intervention
In the intervention group almost all individuals with
diabetes or hypertension were exposed to the mass
education effort. We delivered 20,000 brochures and
5,000 DVDs with psycho-educational content to homes
in the intervention neighborhoods. Each DVD had two
to eight lectures. The DVDs were also broadcast for a
total of 514 days in community venues for a total time
of approximately 4000 hours.
A total of 325 individuals participated in the smallgroup peer support intervention, that is, only 30.8% of
the 1055 participants who were eligible (baseline PHQ-9
or GAD-7 score >5) for this intervention. They were
divided into 28 peer support groups that met a total of
575 times. The mean (sd) attendance by each of these
participants was 17.3 (8.6) times.
A total of 24 individuals received individualized
sessions of PST, that is, only 8.9% of the 269 participants
who were eligible (baseline PHQ-9 or GAD-7 score >10)
for this intervention. In total, 83 individual counseling
sessions were held; the mean (sd) frequency of
counseling sessions for these individuals was 4.3 (2.4)
times.
3.6 Evaluation of the outcome of the intervention
The results of the intervention are shown in Tables
3 and 4. Table 3 compares the continuous outcome
measures, that is, the total scores for the PHQ-9, GAD-7,
and the Physical Component Summary (PCS) and Mental
Component Summary (MCS) scores of the SF-12. In
the control group, the self-reported level of depression
and anxiety became more severe over the 6-month
follow-up, the PCS score did not change significantly,
and the MCS score got worse. Over the same period
in the intervention group, the level of depression did
not change significantly, the level of anxiety improved,
the PCS score did not change significantly, and the MCS
score improved significantly. At both baseline and at the
6-month follow-up assessment the intervention group
had significantly less severe depression, less severe
anxiety, and better PCS and MCS scores than the control
group. After adjusting for the baseline differences of the
measures and for the demographic variables that were
significantly different between the groups at baseline,
at the 6-month follow-up the intervention group still
had significantly less severe depression, significantly less
severe anxiety, and a significantly higher MCS scores
than the control group.
Table 4 compares the dichotomous outcome
measures between the groups. Among the 1239
individuals who completed both assessments in the
control group and the 3039 individuals who completed
both assessments in the intervention group, the
classification of the subtypes of depressive and
anxiety symptoms at baseline was as follows: (a) the
prevalence of mild or more severe depressive symptoms
(PHQ-9 >5) was 17.6% versus 12.5%, respectively;
(b) the prevalence of moderate or severe depressive
symptoms (PHQ-9 >10) was 4.6% versus 5.6%,
respectively; (c) the prevalence of mild or more severe
anxiety symptoms (GAD-7 >5) was 8.1% versus 3.5%,
Table 3. Comparison of mean (sd) results in the intervention group subjects and control group subjects
who completed both the baseline and the 6-month follow-up assessments
control group
scale
n
intervention group
at 6
paired
baseline months
t-test (p)
n
comparison of control and
intervention groups
at
at
at
at 6
paired baseline 6 months 6 months
baseline months
t-test (p) t-test (p) t-test (p) F-test (p)a
PHQ-9
1239
2.39
(3.42)
3.04
(3.44)
5.64
3039
(<0.001)
1.90
(3.17)
1.81
(3.25)
1.30
(0.194)
4.36
(<0.001)
10.81
(<0.001)
9.98
(<0.001)
GAD-7
1239
1.16
(2.36)
1.74
(2.58)
6.67
3039
(<0.001)
0.88
(2.11)
0.73
(1.96)
3.41
(0.001)
3.65
(<0.001)
12.48
(<0.001)
12.85
(<0.001)
SF-12-PCS
1207
44.9
(8.7)
45.1
(8.0)
0.44
(0.664)
2954
46.2
(8.4)
46.0
(8.5)
1.55
(0.121)
4.26
(<0.001)
3.26
(0.001)
1.03
(0.306)
SF-12-MCS 1207
54.4
(8.8)
51.9
(8.5)
8.03
2954
(<0.001)
55.6
(8.3)
56.5
(7.5)
5.28
(<0.001)
3.87
(<0.001)
16.46
(<0.001)
16.13
(<0.001)
PHQ-9, 9-item Patient Health Questionnaire[13]
GAD-7, 7-item General Anxiety Disorder scale[14]
SF-12-PCS, Physical Component Summary score computed by weighting items of the 12-item Short Form Health Survey[15]
SF-12-MCS, Mental Component Summary score computed by weighting items of the 12-item Short Form Health Survey[15]
a
F-test for analysis of covariance that controls for baseline value and for demographic variables that were different at baseline (i.e., occupation,
occurrence of sequelae, and hospital-based treatment in prior 6 months).
• 81 •
Table 4. Comparison of proportions of respondents with mild or or more severe depression or anxiety
(PHQ-9 or GAD-7 total score >5) or moderate or severe depression or anxiety (PHQ-9 or GAD-7 total
score > 10) among intervention group and control group respondents who completed both the
baseline and 6-month follow-up assessments
control group
scale
n
PHQ-9 >5
1239
GAD-7 >5
1239
PHQ-9 >10
1239
GAD-7 >10
1239
PHQ-9 or
GAD-7 >5
1239
PHQ-9 or
1239
GAD-7 >10
self-report
of unstable
1225
blood pressure
control
self-report
of unstable
1210
diabetes
control
intervention group
6
baseline
McNemar
months
test (p)
n (%)
n (%)
218
(17.6)
100
(8.1)
57
(4.6)
17
(1.4)
236
(19.0)
60
(4.8)
322
(26.0)
167
(13.5)
83
(6.7)
22
(1.8)
336
(27.1)
88
(7.1)
32.00
(<0.001)
22.11
(<0.001)
5.30
(0.021)
0.43
(0.511)
28.74
(<0.001)
5.79
(0.016)
---
37
(3.0)
---
---
59
(4.9)
---
n
6
baseline
McNemar
months
test (p)
n (%)
n (%)
comparison of control and
intervention groups
at
at
at
baseline
6 months 6 months
OR
OR
OR
(95% CI)
(95% CI)
(95% CI)a
0.67
0.35
0.36
(0.56-0.80) (0.29-0.41) (0.30-0.43)
0.68
0.34
0.34
(0.53-0.88) (0.27-0.42) (0.27-0.43)
0.75
0.57
0.60
(0.54-1.04) (0.43-0.76) (0.45-0.81)
0.86
0.53
0.60
(0.48-1.54) (0.31-0.93) (0.34-1.07)
0.67
0.34
0.36
(0.57-0.80) (0.29-0.41) (0.30-0.43)
0.74
0.55
0.58
(0.54-1.03) (0.41-0.73) (0.43-0.78)
379
(12.5)
106
(3.5)
171
(5.6)
36
(1.2)
416
(13.7)
111
(3.7)
332
(10.9)
120
(3.9)
151
(5.0)
29
(1.0)
345
(11.4)
122
(4.0)
4.29
(0.043)
0.90
(0.343)
1.38
(0.240)
0.68
(0.410)
9.42
(<0.001)
0.51
(0.474)
3015
---
101
(3.3)
---
---
0.90
(0.61-1.32)
---
2979
---
107
(3.6)
---
---
1.38
(0.99-1.90)
---
3039
3039
3039
3039
3039
3039
PHQ-9, 9-item Patient Health Questionnaire[13]
GAD-7, 7-item General Anxiety Disorder scale[14]
SF-12-PCS, Physical Component Score of 12-item Short Form Health Survey[15]
SF-12-MORCS, Mental Component Score of 12-item Short Form Health Survey[15]
OR, Odds Ratio
95% CI, 95 percent Confidence Interval
a
Odds ratio adjusted for baseline values value and for demographic variables that were different at baseline (i.e., occupation, occurrence of sequelae,
and hospital-based treatment in prior 6 months,).
respectively; (d) the prevalence of moderate or severe
anxiety symptoms (GAD-7 >10) was 1.4% versus 1.2%,
respectively; (e) the prevalence of mild or more severe
depressive or anxiety symptoms (PHQ-9 or GAD-7 >5)
was 19.0% versus 13.7%, respectively; and (f) the
prevalence of moderate or severe depressive or anxiety
symptoms (PHQ-9 or GAD-7 >10) was 4.8% versus
3.7%, respectively. At baseline the prevalence of mild
(or more severe) depressive symptoms, mild anxiety
symptoms, and mild depressive or anxiety symptoms
was significantly greater in the control group than in the
intervention group.
In the control group, the prevalence of mild or
more severe depressive symptoms, mild or more severe
anxiety symptoms, moderate or severe depressive
symptoms, and mild or moderate depressive or
anxiety symptoms increased significantly over the
6-month follow-up period. Over the same period in
the intervention group the prevalence of mild or more
severe depressive symptoms decreased significantly
and the prevalence of mild or more severe depressive
or anxiety symptoms also decreased significantly. The
prevalence of all six measures was significantly lower in
the intervention group than in the control group at the
6-month follow-up assessment. Five of the 6 measures
(with the exception of the prevalence of moderate
or severe anxiety symptoms) remained significantly
different between groups even after adjusting for the
baseline prevalence and for demographic and clinical
variables that were significantly different between the
groups at baseline.
At the 6-month follow-up the self-reported rate of
unstable hypertension and unstable diabetes was not
significantly different between individuals in the control
and intervention groups.
• 82 •
4. Discussion
4.1 Main findings
This 6-month community-based study was a large-scale
effort aimed at assessing the feasibility of reducing
the severity of depressive and anxiety symptoms
of individuals with diabetes or hypertension in an
environment where mental health personnel are
extremely limited. At baseline the prevalence of selfreported mild or more severe depressive or anxiety
symptoms (assessed using the PHQ-9 and the GAD-7)
in 8813 community residents receiving treatment for
diabetes or hypertension was 16% and the prevalence
of moderate or severe depressive or anxiety symptoms
(i.e., clinically significant symptoms) was 4%. We
encountered substantial difficulties in implementing
such a large intervention project (described below),
but the overall outcome – based on the self-report
of participants – indicates that the multi-component
intervention substantially reduced the severity of
both depressive and anxiety symptoms in individuals
receiving routine care for diabetes or hypertension. We
also found that the intervention was associated with an
improvement in the mental health component of quality
of life (assessed by the Mental Component Summary
score of the SF-12), but not in the physical health
component of quality of life (assessed by the Physical
Component Summary score of the SF-12) or in the selfreported rates of uncontrolled diabetes or hypertension.
Our results about changes in depressive and
anxiety symptoms associated with the psychological
intervention (primarily community-based mental
health education campaign) are largely consistent with
results from other countries. The rapid epidemiological
transition (and aging of the population) in high-income
countries and many low- and middle-income countries
is resulting in dramatic increases in the prevalence
of non-communicable diseases such as diabetes and
hypertension, a trend that is particularly evident in
China. One potential approach to reducing the health
burden of such conditions in high-income countries is to
manage the psychological symptoms that often co-exist
with these chronic physical conditions.[21] The results of
studies in this area are not entirely consistent, but the
weight of the evidence supports the value of alleviating
symptoms of depression and anxiety in individuals with
chronic medical conditions.[22,23] Based on these findings,
international practice guidelines, such as those proposed
by the International Diabetes Federation (IDF),[24] stress
the need to address psychological disorders in the
management of individuals with diabetes.
Previous studies in the international and Chinese
literature suggest that psychological interventions can
significantly improve the indicators of somatic health
such as blood pressure[23,25] and blood sugar levels,[23,26]
but the conclusions from systematic reviews of these
studies are inconclusive.[6,27-29] In this study we did not
find differences in the change in the clinical status of
diabetes or hypertension between the intervention
and control groups, but our assessment of the somatic
effects of the intervention were limited to selfreports of having ‘unstable’ hypertension or ‘unstable’
control of blood sugar levels, and to self-reports of the
Physical Component Summary score of the SF-12, so
the study may not have been sensitive to changes in
these physical conditions. Previous studies about the
correlation of objective measures of blood pressure
and blood sugar levels with self-reports of blood
pressure monitoring[30] and self-reports of blood sugar
monitoring[31] show variable results, so basing a decision
about the effectiveness of an intervention on such selfreport measures is probably unwise. At the very least,
future studies need to include assessment of baseline
and post-intervention blood pressure and fasting blood
glucose levels.
Depression, anxiety, and chronic illness all negatively
affect an individuals’ quality of life. Several authors[23,32]
suggest that psychological interventions that alleviate
symptoms of depression or anxiety in individuals with
chronic medical conditions can simultaneously improve
the individuals’ quality of life. The present study found
that our community-based psychological intervention
was associated with improvement in the psychological
component of quality of life (the MCS score for the
SF-12) but not in the somatic component of quality of
life (the PCS score of the SF-12). This result is consistent
with the findings of a systematic review of collaborative
care[21] and with a study on the treatment of depression
in individuals with coronary artery disease.[5]
4.2 Limitations
This study has several major limitations that should be
considered when interpreting the results. We included
community residents registered at three community
health centers (CHCs) in Shanghai with diabetes or
hypertension, but the included CHCs may not be
representative of all CHCs in Shanghai, and, more
importantly, the management rates of hypertension
and diabetes in Shanghai communities is only about
40%,[33] so there may be a selection bias which limits
the generalization of the results. Other factors that
affect the representativeness of the sample on which
the assessment of the outcome of the intervention
was based (i.e., individuals who completed both the
baseline and follow-up evaluations) included: (a)
relatively high dropout rates for both the intervention
group (56%) and the control group (35%); (b) significant
differences in the demographic characteristics, clinical
characteristics, and baseline results for the outcome
variables of interest between those who those who do
and do not complete the study; and (c) restriction of the
outcome assessment for the intervention group to the
19 neighborhoods (out of 34 neighborhoods) where the
intervention was considered most effective. The initial
intention to balance the proportion of participants
with hypertension and diabetes in the intervention
and control groups was not effective: the much higher
proportion of intervention group participants than
control group participants with hypertension (without
comorbid diabetes) who completed the study (70% v.
54%) is particularly concerning because most reports
suggest that hypertension is less likely to be associated
with depressive and anxiety symptoms than diabetes.[12]
Another major problem with the study was the
low participation rate in the small-group peer support
effort (31% of eligible individuals participated) and in
the PST counseling component of the intervention (9%
of eligible individuals participated). Only 349 of the
6897 (5%) individuals in the intervention neighborhoods
who completed the baseline assessment participated
in these components of the intervention, so it is
unlikely that these components of the intervention had
much effect on the overall results; thus the outcome
assessment primarily reflected the outcome of the
mass education campaign. Potential reasons for the low
participation in these components of the intervention
include: (a) patients were invited to participate by the
community clinicians, some of whom were unable
or unwilling to take the time to explain the potential
value of the psychological intervention to the target
recipients; (b) concerns about privacy, confidentiality,
and the stigma of being labeled as ‘mentally ill’ limited
participants’ willingness to join peer support groups;
and (c) the volunteer counselors who provided PST were
unknown to the participants and, moreover, had little
experience in working with elderly patients.
Other limitations of the study include: (a)
assignment to the intervention and control groups
was based on the community health service teams (6
assigned to the intervention group and 5 to the control
group) and this assignment was not done randomly,
so strictly speaking the analysis should be based on
comparing the mean results in these 11 ‘clusters’, not on
the results of all individuals who are in the intervention
and control communities; (b) all the evaluations of
outcome were based on self-completion forms; (c) there
was no clinical assessment of participants to determine
the proportion who meet diagnostic criteria for
depression or anxiety disorders; (d) all the evaluations
were non-blinded; and (e) we did not have data on
blood pressure and fasting blood glucose before and
after the intervention, so it was not possible to assess
the effect of the program on the clinical status of the
participants.
4.3 Significance
We find that clinically significant depressive and
anxiety symptoms are relatively common in community
residents in Shanghai being treated at local CHCs for
diabetes or hypertension. Given the negative effect of
these psychological problems on the quality of life and
prognosis of individuals with these common chronic
physical disorders,[5,6] developing effective strategies
to reduce the prevalence of depressive and anxiety
symptoms in these individuals is an important public
• 83 •
health objective. But the severe lack of mental health
manpower and the stigma associated with receiving
mental health treatment in low- and middle-income
countries (including Shanghai), makes the individualbased psychiatric and psychotherapeutic approaches
employed in high-income countries impractical. As a
first step to address this problem, we implemented
a 6-month multi-component community-based
intervention in 62 neighborhoods in Shanghai that
had a total of 10,164 individuals registered with
hypertension and/or diabetes at local community
health centers. There were several methodological
challenges in the implementation of such a huge project
– selection bias in the evaluation of the outcome, poor
fidelity in the implementation of the intervention,
and lack of objective measures to assess changes in
the clinical status of participants – but the outcome of
the study suggests that the intervention can result in
improvement of both depressive and anxiety symptoms
in individuals with diabetes or hypertension. Further,
more rigorously implemented studies will be needed to
confirm these results, but our results suggest that largescale community-based efforts in settings where mental
health resources are very limited can have beneficial
results.
Acknowledgement
We acknowledge the support by the Changning District
Health and Family Planning Commission of the Shanghai
Municipality, the Changning District Mental Health
Center, the Changning District Xinhua Community
Center, the Changning District Community Center
Health Service Division, the Changning District Huayang
Community Center Health Service Division, the Minhang
District Health and Family Planning Commission of the
Shanghai Municipality, the Minhang Mental Health
Center, the Xinzhuang Government of the Minhang
District, and the Minhang District Xinzhuang Community
Center Health Service Division.
Funding
This study was supported by the Key Population
Psychological Health Service program (GWIII-30; this
is a three-year action plan of the Shanghai public
health system, 2011-2013). The funder is the Shanghai
Municipal Commission of Health and Family Planning.
The funder did not participate in the research design,
implementation, data analysis, or drafting of the
manuscript.
The authors declare no conflict of interest.
Informed consent
Every individual who participated in this study signed a
consent form or provided oral consent at the beginning
of the study.
• 84 •
Ethics approval
The ethics committee of the Shanghai Mental Health
Center approved the study (number: 2013-36).
Authors’ contributions
YH was the principal investigator in charge of the overall
design and analysis of the study, and in the review and
revision of the initial manuscript; Q Zeng prepared
the initial draft of the manuscript and participated in
the design, implementation, and analysis of the study;
ZS participated in the design of the study and was in
charge of the implementation of the intervention; WL,
HT, DM, PL, and XZ were in charge of quality control
for the project; XL and XQ conducted related literature
searches, helped clean the data, and participated in
the quality control of the interventions; Q Zhou was in
charge of the data analysis.
综合心理干预对社区慢性病患者的效果评价：一项源于社区的整群、随机、对照试验
曾庆枝，何燕玲，石振宇，刘威青，陶华，卜时明，缪栋蕾，刘萍，张煊昭，李晓萍，齐雪君，周琴
背景：抑郁与焦虑经常出现在慢性躯体疾病患者中，
通常这会加深这些躯体疾病所造成的损失，但是在中
低等收入国家中这一问题却很少受到关注。
目标：评估非专业临床人员和志愿者进行以社区为基
础的心理干预对缓解慢性躯体疾病患者抑郁和焦虑症
状的疗效。
方法：将共计 10,164 名接受糖尿病或高血压治疗的
上海社区居民任意分配到常规治疗组 (n=2042) 或干
预组 (n=8122)，对干预组的干预包括社区范围的心理
健康教育、同伴支持小组和个人咨询。采用自评患者
健康问卷 (Patient Health Questionnaire, PHQ-9)、广泛
性焦虑量表 (Generalized Anxiety Disorder scale, GAD-7)
和 12 项健康状况调查问卷 (12-item Short-Form Health
Survey, SF-12) 来评定基线和干预 6 个月后的抑郁症状、
焦虑症状和生活质量。
结果：8813 人完成了基线评估，其中 16% 的人有轻度
或较严重的抑郁或焦虑症状（PHQ-9 或 GAD-7>5），
并有 4% 的人伴有中度或重度抑郁或焦虑症状（PHQ-9
或 GAD-7>10）。本研究有效实施了干预内容中的健
康教育部分，但是在符合条件成为同伴支持小组的成
员中仅 31% 的对象接受了干预措施，接受个人咨询
的仅 9%。本研究脱落率较高 (51%)，并且在完成和没
有完成随访评估的人群之间存在显著差异。经过这些
混杂因素的调整后，在完成两项评估的对象中，结果
表明抑郁症状 (F=9.98, p<0.001)、焦虑症状 (F=12.85,
p<0.001) 以及 SF-12 中的心理部分总分 (F=16.13,
p<0.001) 均得到显著改善。然而，自我报告未受控制
的糖尿病或高血压的率没有显著变化。
结论：这些结果支持了以社区为基础的干预措施的可
行性，以降低在精神科人力资源有限的中低等收入国
家中慢性疾病患者抑郁和焦虑症状的严重程度。然而，
在确认该措施广泛大规模实施前还有大量方法学上的
挑战需在未来研究中解决。
关键词：抑郁；焦虑；社区干预；糖尿病；高血压；
社区医疗服务；中国
References
1.
Li D, Lv J, Liu F, Liu P, Yang X, Feng Y, et al. Hypertension
burden and control in mainland China: analysis of
nationwide data 2003-2012. Int J cardiol. 2015; 184: 637644. doi: http://dx.doi.org/10.1016/j.ijcard.2015.03.045
6.
Baumeister H, Hutter N, Bengel J. Psychological and
pharmacological interventions for depression in patients
with diabetes mellitus and depression. Diabet Med. 2014;
31(7): 773-786. doi: http://dx.doi.org/10.1111/dme.12452
2.
Xu Y, Wang L, He J, Bi Y, Li M, Wang T, et al. Prevalence and
control of diabetes in Chinese adults. JAMA. 2013; 310(9):
948-959. doi: http://dx.doi.org/10.1001/jama.2013.168118
7.
3.
Khuwaja AK, Lalani S, Dhanani R, Azam IS, Rafique G, White
F. Anxiety and depression among outpatients with type 2
diabetes: a multi-centre study of prevalence and associated
factors. Diabetol Metab Syndr. 2010; 2: 72. doi: http://
dx.doi.org/10.1186/1758-5996-2-72
Coventry P. Multicondition collaborative care intervention
for people with coronary heart disease and/or diabetes,
depression and poor control of hypertension, blood sugar or
hypercholesterolemia improves disability and quality of life
compared with usual care. Evid based med. 2012; 17(6): e13.
doi: http://dx.doi.org/10.1136/ebmed-2012-100570
8.
Duan S, Xiao J, Zhao S and Zhu X. [Effect of antidepressant
and psychological intervention on the quality of life and
blood pressure in hypertensive patients with depression].
Zhong Nan Da Xue Xue Bao Yi Xue Ban. 2009; 34(4):
j.issn:1672-7347.2009.04.007
9.
Li YJ. [The Situation and Affected Factors of Anxiety and
Depression in The Patients with Hypertension].(Master's
Thesis). Beijing: Beijing University of Chinese Medicine;
2013. Chinese
4.
DeJean D, Giacomini M, Vanstone M, Brundisini F. Patient
experiences of depression and anxiety with chronic disease:
a systematic review and qualitative meta-synthesis. Ont
Health Technol Assess Ser. 2013; 13(16): 1-33
5.
Baumeister H, Hutter N, Bengel J. Psychological and
pharmacological interventions for depression in patients with
coronary artery disease. Cochrane Database Syst Rev. 2011;
9: CD008012. doi: http://dx.doi.org/10.1002/14651858.
CD008012.pub3
• 85 •
10. Katon W, Unutzer J, Wells K, Jones L. Collaborative
depression care: history, evolution and ways to enhance
dissemination and sustainability. Gen Hos Psychiatry.
j.genhosppsych.2010.04.001
22. Whalley B, Thompson DR, Taylor RS. Psychological
interventions for coronary heart disease: Cochrane
systematic review and meta-analysis. Int J Behav Med. 2014;
21(1): 109-121. doi: http://dx.doi.org/10.1007/s12529-0129282-x
11. Simon G. Collaborative care for mood disorders. Curr
Opin Psychiatry. 2009; 22(1): 37-41. doi: http://dx.doi.
org/10.1097/YCO.0b013e328313e3f0
23. Katon WJ, Lin EH, Von Korff M, Ciechanowski P, Ludman EJ,
Young B, et al. Collaborative care for patients with depression
and chronic illnesses. New Engl J Med. 2010; 363(27): 26112620. doi: http://dx.doi.org/10.1056/NEJMoa1003955
12. Long J, Duan G, Tian W, Wang L, Su P, Zhang W, et al.
Hypertension and risk of depression in the elderly: a metaanalysis of prospective cohort studies. J Hum Hypertens.
2015; 29(8): 478-482. Epub 2014 Nov 20. doi: http://dx.doi.
org/10.1038/jhh.2014.112
13. Kroenke K, Spitzer RL, Williams JB. The PHQ-9: validity of a
brief depression severity measure. J Gen Intern Med. 2001;
16(9): 606-613. doi: http://dx.doi.org/10.1046/j.15251497.2001.016009606.x
14. Spitzer RL, Kroenke K, Williams JB, Lowe B. A brief measure
for assessing generalized anxiety disorder: the GAD-7. Arch
Intern Med. 2006; 166(10): 1092-1097. doi: http://dx.doi.
org/10.1001/archinte.166.10.1092
15. Dale J, Williams S, Bowyer V. What is the effect of peer
support on diabetes outcomes in adults? A systematic
review. Diabet Med. 2012; 29(11): 1361-1377. doi: http://
dx.doi.org/10.1111/j.1464-5491.2012.03749.x
16. Hegel M, Areán P. Problem-solvingTreatment for Primary
Care: A Treatment Manual for Project Impact. (Thesis
dissertation). Dartmouth University; 2003
17. Ware JE, Kosinski M, Keller SD. How to Score the SF-12
Physical and Mental Health Summary Scales. 3rd ed. Boston:
The Health Institute, New England Medical Center; 1998
18. Bian CD, He XY, Qian J, Wu WY, Li CB. [Effect of
antidepressant and psychological intervention on the quality
of life and blood pressure in hypertensive patients with
depression]. Tong Ji Da Xue Xue Bao (Yi Xue Ban). 2009;
34(4): 136-140. Chinese. doi: http://dx.chinadoi.cn/10.3321/
j.issn:1672-7347.2009.04.007
24. IDF Clinical Guidelines Task Force. Global Guideline for Type
2 diabetes. Brussels: International Diabetes Federation; 2005
25. Dai L, Wang K, Wang WJ. [Effect of psychological intervention
on anxiety or depression and blood pressure of elderly
patients with hypertension in a community]. Zhong Hua Ji
Bing Kong Zhi Za Zhi. 2010; 14(11): 1126-1128. Chinese
26. Huang XF, Song L, Li TJ, Li JN, Li N, Wu SL. [Effect of health
education and psychosocial intervention on depression in
patients with type 2 diabetes]. Zhongguo Xin Li Wei Sheng Za
Zhi. 2002; 16(3): 149-151. Chinese. doi: http://dx.chinadoi.
cn/10.3321/j.issn:1000-6729.2002.03.002
27. Ontario HQ. Screening and management of depression for
adults with chronic diseases: an evidence-based analysis.
Ont Health Technol Assess Ser. 2013; 13(8): 1-45
28. Atlantis E, Fahey P, Foster J. Collaborative care for comorbid
depression and diabetes: a systematic review and metaanalysis. BMJ Open. 2014; 4: e004706. doi: http://dx.doi.
org/10.1136/bmjopen-2013-004706
29. Fu MM, Dong YJ. [Effect of psychological intervention on
depression symptoms and blood glucose level of patients
with diabetes mellitus in China: a meta-analysis]. Zhongguo
Quan Ke Yi Xue. 2013; 16(4): 436-439. Chinese. doi: http://
dx.chinadoi.cn/10.3969/j.issn.1007-9572.2013.02.025
30. Gee ME, Pickett W, Janssen I, Campbell NR, Birtwhistle
R. Validity of self-reported blood pressure control in
people with hypertension attending a primary care center.
Blood Press Monit. 2014; 19(1): 19-25. doi: http://dx.doi.
org/10.1097/MBP.0000000000000018
19. He XY, Li CB, Qian J, Cui HS, Wu WY. [Reliability and validity
of a generalized anxiety disorder scale in general hospital
outpatients]. Shanghai Arch Psychiatry. 2010; 22(4):
j.issn.1002-0829.2010.04.002
31. Quan C, Talley NJ, Cross S, Jones M, Hammer J, Giles N,
et al. Development and validation of the Diabetes Bowel
Symptom Questionnaire. Aliment Pharmacol Ther. 2003;
17(9): 1179-1187. doi: http://dx.doi.org/10.1046/j.13652036.2003.01553.x
20. Lam CL, Tse EY, Gandek B. Is the standard SF-12 health survey
valid and equivalent for a Chinese population? Qual Life
Res. 2005; 14(2): 539-547. doi: http://dx.doi.org/10.1007/
s11136-004-0704-3
32. Von Korff M, Katon WJ, Lin EH, Ciechanowski P, Peterson D,
Ludman EJ, et al. Functional outcomes of multi-condition
collaborative care and successful ageing: results of
randomised trial. BMJ. 2011; 343: d6612. doi: http://dx.doi.
org/10.1136/bmj.d6612
21. Archer J, Bower P, Gilbody S, Lovell K, Richards D, Gask L, et
al. Collaborative care for depression and anxiety problems.
Cochrane Database Syst Rev. 2012; 10: CD006525. doi:
http://dx.doi.org/10.1002/14651858.CD006525.pub2
33. Wu Y, Zhao YP, Huang XX, Wang JY, Xu HL, Su HL.
[Management mode of urban community public health
services within the family doctor system]. Zhongguo Quan
Ke Yi Xue. 2015; 13: 1504-1509. Chinese.
(received, 2016-03-16; accepted 2016-04-15)
Qingzhi Zeng obtained a master’s degree from the Fudan University School of Public Health in 2006. She
has been working at the Clinical Epidemiology Research Institute of the Shanghai Mental Health Center
and the Mental Health Division of the Shanghai Municipal Center for Disease Control and Prevention
since then. She works in the areas of mental health education and health promotion. Her main research
interests are psychiatric epidemiology, community mental health, and the development and evaluation
of scales related to mental health.
• 86 •
Disability, psychiatric symptoms, and quality of life in infertile
women: a cross-sectional study in Turkey
Hacer SEZGIN1, Cicek HOCAOGLU2,*, Emine Seda GUVENDAG-GUVEN3
Background: Infertility is a major life crisis which can lead to the development of psychiatric symptoms
and negative effects on the quality of life of affected couples, but the magnitude of the effects may vary
depending on cultural expectations.
Aim: We compare the level of psychiatric symptoms, disability, and quality of life in fertile and infertile
women in urban Turkey.
Methods: This cross-sectional study enrolled 100 married women being treated for infertility at the
outpatient department of the Obstetrics and Gynecology Department of the Rize Education and Research
Hospital and a control group of 100 fertile married women. All study participants were evaluated with
a socio-demographic data screening form, the Hospital Anxiety and Depression Scale (HADS), the Brief
Disability Questionnaire (BDQ), and the Short Form Health Survey (SF-36).
Results: The mean anxiety subscale score and depression subscale score of HADS were slightly higher in
the infertile group than in controls, but the differences were not statistically significant. The proportion of
subjects with clinically significant anxiety (i.e., anxiety subscale score of HADS >11) was significantly higher
in infertile women than in fertile women (31% v. 17%, X2=5.37, p=0.020), but the proportion with clinically
significant depressive symptoms (i.e., depression subscale score of HADS >8) was not significantly different
(43% v. 33%, X2=2.12, p=0.145). Self-reported disability over the prior month was significantly worse in the
infertile group than in the controls, and 4 of the 8 subscales of the SF-36 – general health, vitality, social
functioning, and mental health – were significantly worse in the infertile group. Compared to infertile
women who were currently working, infertile women who were not currently working reported less severe
depression and anxiety and better general health, vitality, and mental health.
Conclusions: Married women from urban Turkey seeking treatment for infertility do not have significantly
more severe depressive symptoms than fertile married controls, but they do report greater physical and
psychological disability and a poorer quality of life. The negative effects of infertility were more severe in
infertile women who were employed than in those who were not employed. Larger follow-up studies are
needed to assess the reasons for the differences between these results and those reported in western
countries which usually report a higher prevalence of depression and anxiety in infertile patients.
Keywords: infertility; quality of life; disability; psychiatric symptoms; cross-sectional study; Turkey
1
Department of Family Medicine, Recep Tayyip Erdogan University School of Medicine, Rize, Turkey
Department of Psychiatry, Recep Tayyip Erdogan University School of Medicine, Rize, Turkey
3
Department of Obstetrics and Gynecology, Karadeniz Technical University, School of Medicine, Trabzon, Turkey
2
*correspondence: Dr. Cicek Hocaoglu, Department of Psychiatry, Recep Tayyip Erdogan University School of Medicine, Rize, Turkey.
E-mail: [email protected]
1. Introduction
Infertility, defined as the failure to become pregnant
despite regular sexual intercourse for one year, affects
10-15% of couples in the reproductive age group (1845 years of age).[1] It often results in substantial negative
social and psychological effects for the affected couple,
particularly the woman. [2-4] There are many studies
about the etiology and treatment of infertility[5-7] but
relatively few about the psychological and social effects
of infertility.
One study of 112 women being treated for infertility
in Taiwan[8] reported that 23% met diagnostic criteria
for an anxiety disorder, 17% for major depressive
disorder, and 10% for dysthymic disorder; thus over 40%
had one of these common mental disorders, a much
higher prevalence than the 10% to 12% reported in the
general population. Nationally representative studies of
community-dwelling women in the United States,[9]and
in Finland[10] reported that infertility was associated with
high rates of anxiety symptoms.
Social factors influence attitudes about infertility
and the lived experience of persons who are infertile.
Thus, it is reasonable to expect that the prevalence of
mental disorders in individuals with infertility will vary
cross-culturally. The aim of this study was to compare
the severity of anxiety, depression, and diminished
quality of life between married women from one
urban center in Turkey seeking treatment for infertility
with that of fertile married women from the same
community who are matched for age.
• 87 •
2. Methods
2.1. Participants
As shown in Figure 1, this study enrolled married
women treated in the outpatient clinic of the
Department of Obstetrics and Gynecology of the Rize
Training and Research Hospital who had a diagnosis
of infertility between March and September 2011.
Participants met the following criteria: (a) 18 to 50
years of age; (b) currently married; (c) residents of
Rize; (d) able to read at a level that made it possible
to complete the questionnaires used in the study; (e)
not menopausal; (f) did not have mental retardation,
dementia, a psychotic disorder, or a history of substance
abuse; (g) had not used psychoactive medication in
the prior 3 months; and (h) provided written informed
consent to participate in the study. The control group
were healthy fertile women who were currently married
and residents of Rize; they were identified from among
hospital workers and relatives of the enrolled patients,
matched for age with the identified patients, and
provided written informed consent to participate in the
study.
2.2. Measurements
All participants were administered a comprehensive
demographic data form by the researcher, and selfcompleted three scales: the Turkish versions of the
Hospital Anxiety and Depression Scale (HADS),[11] the
Brief Disability Questionnaire (BDQ),[12] and the Short
Form Health Survey (SF-36).[13]
108 female married outpatients with infertility
treated at the Training and Research Hospital
of Recep Tayyip Erdogan University from
March to September 2011
100 healthy, married, fertile female volunteers
recruited from March to September 2011 and
matched with cases by age
6 refused to participate
102 enrolled patients completed the Hospital
Anxiety and Depression Scale (HADS), the
Brief Disability Questionnaire (BDQ), and the
Short Health Survey Form (SF-36)
100 enrolled controls completed the Hospital
Anxiety and Depression Scale (HADS), the Brief
Disability Questionnaire (BDQ), and the Short
Health Survey Form (SF-36)
2 did not complete testing
100 infertile outpatients completed the study
100 fertile controls completed the study
• 88 •
2.2.1 The Hospital Anxiety and Depression Scale
The Hospital Anxiety and Depression Scale (HADS)[14] is a
14-item scale (7 about anxiety and 7 about depression)
scored on 4-point Likert scales (ranging from 0 to 3)
that assesses the severity of depressive and anxiety
symptoms in the prior week. The total score for each
of the two subscales, respectively) ranges from 0 to 21,
with higher scores representing more severe depression
or anxiety. Based on studies with the Turkish version of
the scale,[11] individuals with scores of 8 or above on the
depression subscale have clinically significant depression
and individuals with scores of 11 or more on the anxiety
subscale have clinically significant anxiety.
2.2.2 The Brief Disability Questionnaire
The Brief Disability Questionnaire (BDQ) is composed of
11 items about physical and social deficits in the prior
month that were originally part of the MOS Short Form
General Health Survey.[15] Items are scored on 3-point
Likert scales (0 to 2), so the range in scores is from 0
to 22 with higher scores representing greater deficits:
scores of 0 to 4 are classified as ‘no deficit’, 5 to 7 as ‘mild
deficit’, 8 to 12 as ‘moderate deficit’, and 13 or higher as
‘severe deficit’. The validity and reliability of the Turkish
version of BDQ have been assessed.[12]
2.2.3 The Short Form Health Survey
The Short Form Health Survey (SF-36) [15] is a selfcompletion scale developed by the Rand Corporation to
assess quality of life. The 36 items are subdivided into
8 subscales that assess physical functioning, physical
role performance, pain, general health, vitality (energy),
social functioning, emotional role-performance, and
mental health. The crude subscale scores are converted
to 0-to-100 point scales with higher scores representing
better health status. The validity and reliability of
the Turkish version of the scale has been assessed
previously.[18]
2.3 Statistical Analysis
Data were assessed using the SPSS v16.0 statistical
package. Demographic variables and the outcomes of
the three clinical self-report scales used in the study in
the infertile and fertile groups were compared using
Chi-square tests for dichotomous variables, MannWhitney U tests for ranked variables, and t-tests for
continuous variables from normal populations. Within
the infertile group, the relationship of the demographic
characteristics of the individuals with the outcomes
of the three scales were assessed using correlation
coefficients (for continuous variables), Chi-square tests,
and the Mann-Whitney U test.
The conduct of this study was approved by the
Clinical Research Ethics Committee in the Faculty of
Medicine at Recep Tayyip Erdogan University.
3. Results
In total, 100 infertile women and 100 healthy volunteers
completed the study. Table 1 compares the demographic
characteristics of the two groups. There were no
significant differences in the level of education or family
income between the infertile and fertile women, in the
proportion who were currently employed, or in the
proportions who reported a personal or family history of
psychiatric treatment. The range in age of individuals in
the infertile group was 21 to 47 and that of individuals
in the control group was 22 to 52. The mean (sd) age
of individuals in the infertile group was 29.7 (5.6) years
and that in the fertile control group was 30.7 (5.5) years
(t=1.27, p=0.204). There was, however, a significant
difference in the duration of marriage between groups:
the infertile group had been married for an average
of 9.3 (6.3) years while the healthy control group had
only been married for an average of 6.4 (3.4) years
(t=4.05, p<0.001). Among the 8 women in the infertile
group with a history of psychiatric illness, 5 had had
major depression, 2 panic disorder, and 1 somatization
disorder; the 11 women in the healthy control group
with a history of a psychiatric disorder included 6 who
had had major depression, 3 with generalized anxiety
disorder, 2 with adjustment disorder, and 1 with
obsessive-compulsive disorder.
Comparison of the anxiety and depression subscale
scores of the HADS, BDQ total scores, and SF-36
subscale scores between the two groups is shown in
Table 2. The mean level of self-reported anxiety and
depressive symptoms over the prior week was not
significantly different between the two groups. However,
the proportion of subjects who had clinically significant
anxiety (i.e., HADS anxiety subscale score >11) was
significantly higher in the infertile group than in the
control group (31% v. 17%, X2=5.37, p=0.020) and the
proportion who had clinically significant depression (i.e.,
HADS depression subscale score >8) was also higher (but
not significantly higher) in the infertile group than in the
control group (43% v. 33%, X2=2.12, p=0.145).
The severity of self-reported disability was
significantly greater among infertile patients than among
the fertile controls. The proportion of respondents in
the infertile group classified as ‘no disability’, ‘mild’
disability’, ‘moderate disability’ and ‘severe disability’
were 5%, 15%, 63%, and 17%, respectively; the
corresponding proportions in the fertile control group
were 39%, 39%, 20% and 2%, respectively. (Z-value
for the Mann-Whitney rank test=7.82, p<0.001).
Comparison of the scores of the various measures
assessed by the SF-36 show that 4 of the 8 subscales –
general health, vitality, social functioning, and mental
health – were significantly worse in the infertile group.
Table 3 shows the association between different
demographic characteristics of the infertile patients and
the severity of their depressive and anxiety symptoms,
their self-reported level of disability, and their scores
on the four SF-36 subscales in which the infertile
• 89 •
Table 1. Comparison of socio-demographic and clinical characteristics of infertile female patients and healthy,
fertile controls
infertile patients
(n=100)
n (%)
healthy controls
(n=100)
n (%)
1 (1%)
4 (4%)
primary school
38 (38%)
36 (36%)
middle school
16 (16%)
15 (15%)
high school
27 (27%)
32 (32%)
18 (18%)
13 (13%)
<1000 TL
30 (33.3%)
42 (42.4%)
1001-2000 TL
28 (31.1%)
35 (35.4%)
2001-3000 TL
21 (23.3%)
15 (15.2%)
≥3001 TL
11 (12.2%)
7 (7.1%)
Currently employed
74 (74%)
83 (83%)
X2=5.16 (0.473)
8 (8%)
15 (15%)
X2=2.13 (0.144)
12 (12%)
11 (11%)
X2=0.04 (0.835)
characteristic
statistic (p-value)
Educational status
illiterate
university
Family income (Turkish lira, TL)
b
History of psychiatric illness
Family history of psychiatric illness
a
b
Za=0.48 (0.631)
Za=1.77 (0.077)
Z-value for Mann-Whitney U test
In September 2011, 1.78 Turkish lira were equivalent to 1 $US; 10 patients in the infertile group and 1 in the control group did not
provide income data
Table 2. Mean (sd) scores from the Hospital Anxiety and Depression Scale (HADS), the Brief Disability
Questionnaire (BDQ), and the Short Form Health Survey (SF-36) of 100 infertile female patients and
100 fertile controls from Turkey
infertile patients
fertile controls
t-test
p-value
HADS anxiety subscale
8.2 (4.3)
7.3 (4.1)
1.51
0.131
HADS depression subscale
6.6 (4.1)
6.3 (3.4)
0.56
0.574
BDQ
9.1 (2.8)
5.4 (3.2)
8.70
<0.001
physical functioning
78.3 (19.9)
80.3 (15.6)
0.79
0.430
physical role performance
58.5 (40.0)
49.7 (38.3)
1.59
0.114
pain
63.9 (20.4)
60.2 (17.6)
1.37
0.171
general health
47.4 (22.3)
60.5 (18.2)
4.55
<0.001
vitality (energy)
41.3 (22.9)
52.4 (17.8)
3.82
<0.001
social functioning
56.5 (23.2)
67.8 (21.3)
3.59
<0.001
emotional role performance
50.6 (38.3)
52.6 (39.6)
0.36
0.717
mental health
55.2 (23.2)
61.4 (20.4)
2.01
0.046
SF-36 subscales
• 90 •
Table 3. Association of demographic variables and scores of the Hospital Anxiety and Depression Scales
(HAD-D, HAD-A), the Brief Disability Questionnaire (BDQ), and three subscale scores of the Short
Form Health Survey (SF-36) in 100 infertile female outpatients in Turkey
BDQ
SF-36
general
health
subscale
SF-36
vitality
subscale
SF-36
social
function
subscale
SF-36
mental
health
subscale
0.004
(0.969)
-0.25
(0.013)
0.07
(0.465)
0.10
(0.307)
0.07
(0.744)
0.08
(0.387)
-0.24
(0.005)
-0.27
(0.012)
0.15
(0.130)
0.10
(0.293)
0.08
(0.387)
-0.96
(0.001)
0.07
(0.473)
-0.22
(0.025)
-0.21
(0.031)
0.07
(0.485)
0.20
(0.046)
0.12
(0.243)
0.09
(0.361)
0.14
(0.161)
<5 years (n=58)
6.0 (4.2)
7.9 (4.3)
9.7 (2.8) 48.1 (22.7) 50.9 (23.7) 55.1 (25.1) 56.1 (24.3)
5+ years (n=42)
7.5 (3.9)
8.6 (4.3)
8.4 (2.7) 46.5 (22.0) 51.9 (21.9) 58.3 (20.5) 54.0 (21.9)
t-test
(p-value)
1.72
(0.088)
0.82
(0.410)
2.22
(0.029)
currently employed (n=74)
7.5 (4.0)
9.0 (4.1)
9.0 (2.6) 43.3 (20.7) 47.2 (22.0) 53.8 (22.0) 50.0 (22.8)
not currently employed (n=26)
4.1 (3.2)
5.8 (4.0)
9.6 (3.4) 59.3 (22.8) 63.0 (21.6) 63.9 (25.3) 70.1 (17.6)
t-test
(p-value)
3.90
(<0.001)
3.47
(0.001)
1.03
(0. 304)
yes (n=8)
7.0 (5.8)
9.8 (5.3) 10.8 (2.8) 49.6 (24.4) 45.6 (27.1) 43.7 (21.1) 48.0 (30.3)
no (n=92)
6.6 (4.0)
8.0 (4.2)
9.0 (2.8) 47.2 (22.3) 51.8 (22.6) 57.6 (23.2) 55.9 (22.6)
t-test
(p-value)
0.23
(0.816)
1.12
(0.263)
1.34
(0.181)
yes (n=12)
6.2 (3.7)
9.0 (4.6)
8.7 (2.8) 54.3 (22.1) 48.3 (19.9) 61.4 (22.2) 53.0 (19.0)
no (n=88)
6.7 (4.2)
8.1 (4.3)
9.2 (2.9) 46.5 (22.3) 51.7 (23.3) 55.8 (23.4) 55.5 (23.8)
t-test
(p-value)
0.37
(0.710)
0.66
(0.509)
0.56
(0.575)
HADS
depression
subscale
HADS
anxiety
subscale
age,
Pearson r (p-value)
0.09
(0.343)
level of education,
Spearman r (p-value)
monthly income,
Spearman r (p-value)
YEARS OF MARRIAGE, mean (sd)
0.36
(0.718)
0.20
(0.838)
0.66
(0.505)
0.43
(0.667)
EMPLOYMENT STATUS, mean (sd)
3.28
(0.001)
3.17
(0.002)
1.92
(0.058)
4.07
(<0.001)
HISTORY OF PSYCHIATRIC
ILLNESS, mean (sd)
0.28
(0.779)
0.73
(0.464)
1.62
(0.106)
0.92
(0.359)
FAMILY HISTORY OF
PSYCHIATRIC ILLNESS, mean (sd)
patients were functioning at significantly lower levels
than controls. There were several significant findings.
AGE: somewhat unexpectedly, within this group of
infertile women, self-reported disability decreased with
age. EDUCATION: higher education was significantly
associated with decreased self-reported depression and
anxiety, and poorer self-reported social functioning.
INCOME: higher family income was associated with less
severe self-reported depression and anxiety, and better
self-reported general health. DURATION OF MARRIAGE:
1.13
(0.260)
0.48
(0.629)
0.78
(0.434)
0.36
(0.720)
infertile women married for less than 5 years reported
significantly greater disability over the prior month than
infertile women married for 5 years or more. CURRENT
EMPLOYMENT: compared to employed infertile women,
unemployed infertile women had less severe depressive
and anxiety symptoms and reported better general
health, vitality, and mental health. Neither a HISTORY
OF PSYCHIATRIC ILLNESS nor a FAMILY HISTORY OF
PSYCHIATRIC ILLNESS were significantly related to any
of the outcome variables.
4. Discussion
4.1 Main findings
Both self-report depressive symptoms and self-report
anxiety symptoms on the HADS were more severe
in infertile women than in fertile women, but the
difference was not statistically significant for depressive
symptoms and only statistically significant for anxiety
symptoms when results were dichotomized into
those with and without ‘clinically significant anxiety’.
Infertile women reported greater disability on the BDQ
and poorer functioning on 4 of the 8 components of
quality of life assessed by the SF-36. We also found that
compared to infertile women who were not employed,
those that were employed reported more severe
symptoms of depression and anxiety, greater disability,
and poorer quality of life.
In Turkey, infertile women who are not able to
bear children are marginalized in the society and
often harshly criticized by their husbands and inlaws. This environment would reasonably be expected
to negatively affect the emotional status of infertile
women, and, thus, lead to an increased prevalence
of common mental disorders, such as depression or
anxiety. Most international studies[8,9,16-19] support this
hypothesized causal link between a chronic psychosocial
stressor and emotional dysregulation: they report a
significantly higher severity of depressive and anxiety
symptoms and a significantly higher prevalence of
depressive and anxiety disorders among infertile
women than among fertile women. There are, however,
exceptions: similar to the results of the current study,
two previous studies from Turkey [20,21] reported no
significant difference in the level of depression and
anxiety between infertile and fertile women. Previous
reports have also had different findings about the
association of age and the severity of depression and
anxiety symptoms in infertile women; some studies
confirm our finding of no relationship, [22,23] while
other studies[17,19,20] report that depressive and anxiety
symptoms increase with age. The reason for these
differences are unknown, but the possible explanations
include (a) high levels of depression and anxiety in all
married Turkish women regardless of fertility status; (b)
cross-cultural differences in the mechanism via which
social stressors lead to emotional disturbances; and (c)
methodological limitations of the study,
Several studies have reported on the quality of life
among infertile women.[24-35] Similar to our findings,
most of the case control studies report substantially
decreased quality of life among infertile women in
several of the quality of life subscales.[31] However, unlike
other studies, we did not find that decreased quality of
life among infertile women was closely associated with
increased symptoms of depression.[36-38] Thus the quality
of life changes in our infertile patients in Turkey were
not directly related to changes in the severity of their
psychological symptoms.
Our results related to self-reported disability in the
month prior to the interview were quite robust. Both
• 91 •
the mean score to the BDQ and the ranked classification
of the results of the BDQ found that the infertile
patient group reported significantly greater impairment
than that reported by women of the same age and
marital status who were not infertile. In the absence
of differences in the level of depressive and anxiety
symptoms between the groups, this suggests that social
discrimination of women in Turkey who cannot fulfil this
expected role directly affects their functioning. To our
knowledge, no previous study has reported the level of
disability among infertile subjects.
The reasons for the more prominent depressive
and anxiety symptoms and greater impairment in the
quality of life among employed women who are infertile
compared to that in unemployed women who are
infertile are unknown. Presumably this is related to the
greater exposure employed women who are infertile
have to social disapproval than unemployed women
(who primarily work in the home as housewives), but
further qualitative studies will be needed to clarify this
issue.
4.2. Limitations
This study has several limitations. (a) The cross-sectional
nature of the study made it impossible to identify
causal relationships between infertility and the various
psychological, functional, and quality of life measures
assessed. (b) All measures employed were selfrated, so different types of reporting biases may have
affected the results. (c) There was no formal diagnosis
made of the patients or controls so the proportion
that had psychological disorders that were severe
enough to merit psychiatric intervention was unknown.
(d) The sample was selected from married women
with infertility being treated at an urban outpatient
department, so the results may not be generalizable to
all infertile women. (e) Sexual dysfunction, a common
problem in infertile couples, was not considered among
the eight aspects of quality of life assessed by the SF-36.
(f) Several factors that may affect the psychosocial
effects of infertility (e.g., duration of infertility, use of
different fertility treatments, etc.) were not considered.
Finally, (g) the sample of infertile patients was not
large enough to employ multivariate linear regression
analyses (or other multivariate techniques) to assess
the relative importance of potential demographic and
clinical treatment determinants of depression, anxiety,
perceived disability, or quality of life.
4.3 Importance
This study found that the self-reported level of disability
and levels of several measures of the quality of life of
infertile married women in Turkey, particularly those
who are currently employed, are significantly lower
than those of fertile married women. However, the selfreported level of depressive and anxiety symptoms
was not different between infertile and fertile women.
This disconnect between psychological symptoms,
functioning, and quality of life suggests that western
• 92 •
assumptions about the causal relation of major
psychosocial stressors (such as infertility) to common
mental disorders may need to be adjusted when
considering non-western cultures, where the meaning
and psychological valence of specific types of stressors
can be quite different. Only a minority of infertile
participants had clinically significant depression (43%)
or clinically significant anxiety (33%), so psychosocial
interventions for infertile women should focus on
social support and place somewhat less emphasis on
psychiatric treatment. However, this is a small crosssectional study in one urban clinic in Turkey, so larger
studies that enroll a broader spectrum of infertile
patients and that follow them over time are needed to
confirm the relevance of these findings.
Funding
This study received no financial support.
The authors report no conflict of interest related to this
manuscript.
Ethical review
The study protocol was approved by the Ethics
Committee of the Faculty of Medicine, University of
Recep Tayyip Erdogan, Rize, Turkey. (date of approval:
25.02.2011; number: 2011/6)
Informed consent
Written informed consent was obtained from all
participants.
HS and CH participated in the design of the study,
in data collection, and drafted the manuscript. CH
performed the statistical analysis and critically reviewed
the manuscript. ESGG carried out the clinical diagnosis
and critically reviewed the manuscript. All authors read
and approved the final manuscript.
不育妇女的功能障碍、精神病症状和生活质量：一项来自土耳其横断面研究
Sezgin H, Hocaoglu C, Guvendag-Guven ES
背景：不孕不育是一种重大的生活危机，它可以导
致精神病症状的发展并且对夫妻的生活质量产生负
面影响，但其影响程度可能取决于文化背景。
目标：我们比较了土耳其城市中生育妇女和不孕妇
女的精神病症状程度、功能障碍水平和生活质量。
方法：该横断面研究纳入了 100 名在里泽教育和
研究医院的妇产科门诊治疗不孕不育的已婚女性
和 100 名已婚已育的妇女作为对照组。对所有参与
者均采用社会人口信息筛查表、医院焦虑抑郁量表
(Hospital Anxiety and Depression Scale, HADS)、简单功
能障碍问卷 (Brief Disability Questionnaire, BDQ) 和健
康状况问卷 (Short Form Health Survey , SF-36) 进行评
估。
结果：不育女性的平均焦虑分量表得分和抑郁分量
表得分稍高于对照组，但差异无统计学意义。不
孕组妇女中有显著临床焦虑症状的比例（即焦虑
分量表得分 > 11）显著高于育龄妇女 (31% v. 17%,
X2=5.37, p=0.020)，但有显著临床抑郁症状的比例（即
抑郁分量表评分 HADS > 8）在两组间没有显著性差
异 (43% v. 33%, X2=2.12, p=0.145)。不育女性自我报告
前一个月的功能障碍显著比对照组严重，并且不育
女性在 SF-36 的 8 个分量表中 4 个（一般健康、活力、
社会功能和心理健康）显著差于对照组。与目前工
作的不育女性相比，目前没有工作的女性不育患者
报告的抑郁和焦虑程度较轻，且一般健康状况、活
力和心理健康状况较好。
结论：未发现土耳其城市地区中寻求治疗的不孕不
育已婚女性并比已婚已育妇女有更严重的抑郁症状，
但他们确实报告有较大的躯体和心理障碍并且生活
质量较差。不孕不育的负面影响对在职不孕女性妇
女比无业的不孕妇女更严重。西方国家这通常报告
不孕患者抑郁和焦虑的患病率更高，我们需要更大
规模的随访研究以评估这些结果与西方国家报告的
结果不同的原因。
关键词：不育；生活质量；功能障碍；精神病症状；
横断面研究；土耳其
References
1. Mosher WD, Pratt WF. Fecundity and infertility in the United
States: incidence and trends. Fertil Steril. 1991; 56(2): 192-193
2. Kraft AD, Palombo J, Mitchell D, Dean C, Meyers S, Schmidt
AW. The psychological dimensions of infertility. Am J
Orthopsychiatry. 1980; 50(4): 618-628
3. Sadock BJ, Sadock VA. Synopsis of Psychiatry. 9th ed.
Philadelphia: Lippincott Williams & Wilkins; 2003. p. 60-65
4. Raphael-Leff J. Psychotherapy during the reproductive years.
In Gabbard GO, Beck JS, Holmes J, editors. Oxford Textbook of
Psychotherapy. New York: Oxford University Press; 2005. p.
367-379
5. Nahar P, Richters A. Suffering of childless women in
Bangladesh: the intersection of social identities of gender
and class. Anthropol Med. 2011; 18(3): 327–338. doi: http://
dx.doi.org/10.1080/13648470.2011.615911
6. Onat G, Kızılkaya Beji N. Effects of infertility on gender
differences in marital relationship and quality of life: a case
control study of Turkish couples. Eur J Obst Gynecol Reprod
Biol. 2012; 165(2): 243-248. doi: http://dx.doi.org/10.1016/
j.ejogrb.2012.07.033
7. Mahlstedt PP. The psychological component of infertility. Fertil
Steril. 1985; 43(3): 335-346
8. Chen TH, Chang SP, Tsai CF, Juang KD. Prevalence of depressive
and anxiety disorders in an assisted reproductive technique
clinic. Hum Reprod. 2004; 19(10): 2313-2318. doi: http://
dx.doi.org/10.1093/humrep/deh414
9. King RB. Subfecundity and anxiety in a nationally
representative sample. Soc Sci Med. 2003; 56(4): 739-741.
doi: http://dx.doi.org/10.1016/S0277-9536(02)00069-2
10. Klemetti R, Raitanen J, Sihvo S, Saarni S, Koponen P. Infertility,
mental disorders and well-being: a nationwide survey. Acta
Obstet Gynecol Scand. 2010; 89(5): 677-682. doi: http://
dx.doi.org/10.3109/00016341003623746
11. Aydemir O, Guvenir T, Kuey L, Kultur S. [Reliability and validity
of the Turkish version of the Hospital Anxiety and Depression
Scale]. Turk Psikiyatri Derg. 1997; 8(3): 280-287. Turkish
12. Kaplan I. [The relationship between mental disorders and
disability in patients admitted to the semi-rural health
centers]. Turk Psikiyatri Derg. 1995; 6(2): 169-179. Turkish
13. Koçyigit H, Aydemir O, Fisek G, Olmez N, Memis A. [The
reliability and validity of the Turkish version of Short Form-36
(SF-36)]. İlaç ve Tedavi Dergisi. 1999; 12(3): 102-106. Turkish
14. Aydemir O, Koroglu E. [Clinical scales used in psychiatry].
Hekimler Yayın Birliği. 2006; 138-139: 346-347. Turkish
15. Stewart AL, Hays RD, Ware JE Jr. The MOS Short-Form General
Health Survey: reliability and validity in a patient population.
Med Care. 1988; 26(7): 724-735
16. Anderson KM, Sharpe M, Rattray A, Irvine DS. Distress and
concerns in couples referred to a specialist infertility clinic.
J Psychosom Res. 2003; 54(4): 353-355. doi: http://dx.doi.
org/10.1016/S0022-3999(02)00398-7
17. Domar AD, Zuttermeister PC, Seibel M, Benson H.
Psychological improvement in infertile women after
behavioral treatment: a replication. Fertil Steril. 1992; 58(1):
144-147
18. Lukse MP, Vacc NA. Grief, depression and coping in women
undergoing infertility treatment. Obstet Gynecol. 1999; 93(2):
245-251
19. Drosdzol A, Skrzypulec V. Depression and anxiety among
Polish infertile couples-an evaluative prevalence study. J
Psychosom Obstet Gynaecol. 2009; 30(1): 11-20. doi: http://
dx.doi.org/10.1080/01674820902830276
• 93 •
20. Guz H, Ozkan A, Sarısoy G, Yanik F, Yanik A. Psychiatric
symptoms in Turkish infertile women. J Psychosom Obstet
Gynaecol. 2003; 24(4): 267-271
21. Gulseren L, Cetinay P, Tokatlıoglu B, Sarıkaya OO, Gulseren
S, Kurt S. Depression and anxiety levels in infertile Turkish
women. J Reprod Med. 2006; 51(5): 421-426
22. Ashkani H, Akbari A, Heydari ST. Epidemiology of depression
among infertile and fertile couples in Shiraz, Southern Iran.
Indian J Med Sci. 2006; 60(10): 399-406.
23. Beutel M, Kupfer J, Kirchmeyer P, Kehde S, Kohn FM,
Schroeder-Printzen I. Treatment related stresses and
depression in couples undergoing assisted reproductive
treatment by IVF or ICSI. Andrologia. 1999; 31(1): 27-35. doi:
http://dx.doi.org/10.1111/j.1439-0272.1999.tb02839.x
24. Heredia M, Tenías JM, Rocio R, Amparo F, Calleja
MA, Valenzuela JC. Quality of life and predictive factors in
patients undergoing assisted reproduction techniques. Eur
J Obstet Gynecol Reprod Biol. 2013; 167(2): 176-180. doi:
http://dx.doi.org/10.1016/j.ejogrb.2012.12.011
25. Monga M, Bogdan A, Katz SE, Stein M, Ganiats T. Impact of
infertility on quality of life, marital adjustment and sexual
function. Urology. 2004; 63(1): 126-130. doi: http://dx.doi.
org/10.1016/j.urology.2003.09.015
26. Fekkes M, Buitendijk SE, Verrips GH, Braat DD, Brewaeys
AM, Dolfing JG, et al. Health-related quality of life in relation
to gender and age in couples planning IVF treatment.
Hum Reprod. 2003; 18(7): 1536-1543. doi: http://dx.doi.
org/10.1093/humrep/deg276
27. Hassanin IM, Abd-El-Raheem T, Shahin AY. Primary infertility
and health-related quality of life in Upper Egypt. Int J
Gynecol Obstet. 2010; 110(2): 118-121. doi: http://dx.doi.
org/10.1016/j.ijgo.2010.02.015
28. Abbey A, Andrews FM, Halman LJ. Provision and receipt
of social support and disregard: what is their impact on
the marital life quality of infertile and fertile couples? J
Personality Soc Psychol. 1995; 68(3): 455-469. doi: http://
dx.doi.org/10.1037/0022-3514.68.3.455
29. Andrews FM, Abbey A, Halman LJ. Is fertility problem stress
different? The dynamics of stress in fertile and infertile
couples. Fertil Steril. 1992; 57(6): 1247-1253
30. Andrews FM, Abbey A, Halman LJ. Stress from infertility,
marriage factors, and subjective well-being of wives and
husbands. J Health Soc Behav. 1991; 32(3): 238-253
31. Ragni G, Mosconi P, Baldini MP. Health-related quality
of life and need for IVF in 1000 Italian infertile couples.
Hum Reprod. 2005; 20(5): 1286-1291. doi: http://dx.doi.
org/10.1093/humrep/deh788
32. Weaver SM, Clifford E, Douglas MH, Robinson J. Psychosocial
adjustment to unsuccessful IVF and GIFT treatment.
Patient Educ Couns. 1997; 31(1): 7-18. doi: http://dx.doi.
org/10.1016/S0738-3991(97)01005-7
33. Hearn MT, Yuzpe AA, Brown SE. Psychological characteristics
of in vitro fertilization participants. Am J Obstet Gynecol.
1987; 156(1): 269-274
34. Onat G, Kizilkaya Beji N. Effects of infertility on gender
differences in marital relationship and quality of life: a casecontrol study of Turkish couples. Eur J Obstet Gynecol Reprod
Biol. 2012; 165(2): 243-248. doi: http://dx.doi.org/10.1016/
j.ejogrb.2012.07.033
• 94 •
35. Lau JT, Wang Q, Cheng Y, Kim JH, Yang X, Tsui HY. Infertilityrelated perceptions and responses and their associations
with quality of life among rural Chinese infertile couples. J
Sex Marital Ther. 2008; 34(3): 248-267. doi: http://dx.doi.
org/10.1080/00926230701866117
37. Mosalanejad L, Abdolahifard K, Jahromi MG.
Therapeutic vaccines: hope therapy and its effects
on psychiatric symptoms among infertile women. Glob
J Health Sci. 2013; 6(1): 192-200. doi: http://dx.doi.
org/10.5539/gjhs.v6n1p192
36. Smith JF, Walsh TJ, Shindel AF. Sexual, marital and social
impact of a man’s perceived infertility diagnosis. J Sex Med.
2009; 6(9): 2505-2515. doi: http://dx.doi.org/10.1111/j.17436109.2009.01383.x
38. Carter J, Applegarth L, Josephs L, Grill E. A cross-sectional
cohort study of infertile women awaiting oocyte donation:
the emotional, sexual, and quality-of-life impact. Fertil
Steril. 2011; 95(2): 711-6.e1. doi: http://dx.doi.org/10.1016/
j.fertnstert.2010.10.004
(received, 2016-02-04, accepted, 2016-02-20)
Dr. Hacer Sezgin obtained a medical degree in 2005 from Karadeniz Technical University and received
postgraduate training in family medicine between 2010 and 2013 at the Department of Family
Medicine at the Medical School of Recep Tayyip Erdogan University in Rize, Turkey. She is currently
a specialist physician in the Department of Family Medicine at Çayırli State Hospital in Erzincan,
Turkey. Her research interests are female infertility and its psychological impact, polycystic ovarian
syndrome, diabetes mellitus, insulin resistance, and Hashimoto thyroiditis.
• 95 •
Clinical investigation of speech signal features among patients
with schizophrenia
Jing ZHANG1,2, Zhongde PAN1,3,4, Chao GUI5, Jie ZHU5,*, Donghong CUI1,3,4,*
Background: A new area of interest in the search for biomarkers for schizophrenia is the study of the acoustic
parameters of speech called 'speech signal features'. Several of these features have been shown to be
related to emotional responsiveness, a characteristic that is notably restricted in patients with schizophrenia,
particularly those with prominent negative symptoms.
Aim: Assess the relationship of selected acoustic parameters of speech to the severity of clinical symptoms
in patients with chronic schizophrenia and compare these characteristics between patients and matched
healthy controls.
Methods: Ten speech signal features – six prosody features, formant bandwidth and amplitude, and two
spectral features – were assessed using 15-minute speech samples obtained by smartphone from 26
inpatients with chronic schizophrenia (at enrollment and 1 week later) and from 30 healthy controls (at
enrollment only). Clinical symptoms of the patients were also assessed at baseline and 1 week later using the
Positive and Negative Syndrome Scale, the Scale for the Assessment of Negative Symptoms, and the Clinical
Global Impression-Schizophrenia scale.
Results: In the patient group the symptoms were stable over the 1-week interval and the 1-week test-retest
reliability of the 10 speech features was good (intraclass correlation coefficients [ICC] ranging from 0.55 to
0.88). Comparison of the speech features between patients and controls found no significant differences in
the six prosody features or in the formant bandwidth and amplitude features, but the two spectral features
were different: the Mel-frequency cepstral coefficient (MFCC) scores were significantly lower in the patient
group than in the control group, and the linear prediction coding (LPC) scores were significantly higher in the
patient group than in the control group. Within the patient group, 10 of the 170 associations between the
10 speech features considered and the 17 clinical parameters considered were statistically significant at the
p<0.05 level.
Conclusions: This study provides some support for the potential value of speech signal features as indicators
(i.e., biomarkers) of the severity of negative symptoms in schizophrenia, but more detailed studies using
larger samples of more diverse patients that are followed over time will be needed before the potential
utility of such acoustic parameters of speech can be fully assessed.
Keywords: schizophrenia; speech; speech signal features; biomarkers; negative symptoms; China
[Shanghai Arch Psychiatry. 2016, 28(2): 95-102. doi: http://dx.doi.org/10.11919/j.issn.1002-0829.216025]
1
Jiading District Mental Health Center, Shanghai, China
3
Shanghai Key Laboratory of Psychotic Disorders, Shanghai Mental Health Center, Shanghai Jiao Tong University School of Medicine, Shanghai, China
4
Key Laboratory of Translational Psychiatry, Shanghai Jiao Tong University School of Medicine, Shanghai, China
5
School of Electronic Information and Electrical Engineering, Shanghai Jiao Tong University, Shanghai, China
2
*co-corresponding authors: Dr. Donghong Cui, Shanghai Key Laboratory of Psychotic Disorders, Shanghai Mental Health Center, Shanghai Jiao Tong
University School of Medicine, 3210 Humin Road, Shanghai 201108, China. E-mail: [email protected]; Jie Zhu, School of Electronic Information and
Electrical Engineering, Shanghai Jiao Tong University, 800 Dongchuan Road, Shanghai, China. E-mail: [email protected]
• 96 •
1. Introduction
Schizophrenia is a complex mental disorder caused
by multiple factors including heredity, development,
and environment.[1] The fifth edition of the Diagnostic
and Statistical Manual of Mental Disorders (DSM-5)[2]
lists the following five prominent psychopathological
characteristics of the disorder: delusions, hallucinations,
disorganized speech, grossly disorganized or catatonic
behavior, and negative symptoms. Negative symptoms
include emotional blunting, poverty of speech, avolition,
an inability to experience pleasure, and the lack of desire
to form relationships. Present methods for determining
the diagnosis and for assessing the effectiveness of
treatment primarily rely on the subjective judgment of
the clinician who uses information provided by family
members, the mental status examination, and various
clinical symptom scales. In the absence of objective
measures and the frequent uncooperativeness of
patients – particularly those with prominent negative
symptoms – assessing the severity and course of the
illness is often challenging for clinicians. To address
this fundamental problem, psychiatric researchers are
actively searching for objective biomarkers that can be
used both in the diagnosis of the condition and in the
monitoring of the clinical progress of the disorder.
Fluctuations in speech that parallel patients'
physio-psychological state might be suitable candidates
as biomarkers for schizophrenia. Studies of signal
processing and artificial intelligence find that the
features of speech signals can contain substantial emotional
information.[3,4] Changes of emotions and the range and
variability of emotions can be quantified by changes in
speech parameters, particularly by changes in prosody
– that is, the vocal pitch (fundamental frequency),
loudness (acoustic intensity), and rhythm (phoneme
and syllable duration) of speech. For example, when
a person is in an angry state, changes in physiological
characteristics (e.g., increased heart rate, elevated
skin voltage, and elevated blood pressure) are often
associated with changes in the rate, volume, and tone
of speech.
There is considerable interest in developing
methods for extracting the acoustic parameters which
reflect emotions from speech samples and in assessing
the relationship of these parameters to emotionally
restrictive states, such as the negative symptoms of
schizophrenia. Identification of the emotional content
of speech signals is primarily accomplished by two
processes: first, the features of the speech signals are
extracted from speech samples and then judgments
are made about the emotional content of the identified
features based on pre-existing models. The quality
of the extraction process largely determines the
functional quality of the speech identification system.[5,6]
Studies about speech identification generally start
by investigating the prosodic features and acoustic
characteristics of speech content, focusing on the
features which are directly relevant to the emotional
characteristics of speech.[7]
Patients with schizophrenia who have prominent
negative symptoms such as emotional blunting and
poverty of speech may be particularly prone to having
restricted emotional content in their speech content.
This can directly limit their social functioning and make
it difficult for clinicians to detect changes in their clinical
status over the course of their illness. Identification of
the specific speech abnormalities in such patients could
both help in monitoring the course of the illness and
potentially be used to develop targeted interventions
for patients with prominent negative symptoms.
Several researchers[8-11] have reported relationships
between specific phonetic parameters and the negative
symptoms and impaired emotional perception of
schizophrenia, but the results to date are far from
robust.
The current study assessed the characteristics of
the speech signals of patients with schizophrenia with
prominent negative symptoms, considered the association between these features and the severity of different types of negative symptoms, and compared the
speech signal features in these patients with those in
healthy control subjects.
2. Methods
2.1 Participants
The patient group consisted of patients with
schizophrenia who were inpatients at the Shanghai
Mental Health Center from September 2013 to
December 2015. The inclusion criteria were as follows:
(a) aged from 18 to 65 years; (b) met the diagnostic
criteria of schizophrenia specified in DSM-5 [2] as
assessed by a psychiatrist using the Mini-International
Neuropsychiatric Interview (M.I.N.I. 6.0).[12] (c) had
prominent negative symptoms of schizophrenia; (d)
a minimum of duration of illness of two years; (e) no
co-morbid psychiatric or substance abuse disorder;
(f) no evidence of severe impulsivity; (g) not using
antipsychotic medication that could impair speech; and
(h) both the patient and the patient’s family member
provided written informed consent to participate in
the study, including the use of smartphones to record
speech.
We recruited volunteers from the community by
advertisement as healthy controls. Volunteers were
similar for patients in age and duration of education,
underwent a through psychiatric exam (using the
M.I.N.I. 6.0) and physical exam. Inclusion criteria for
controls were as follows (a) 18 to 65 years of age; (b)
Han Chinese ethnicity; (c) no current or past physiopsychological, substance abuse, or serious neurological
disorder; (d) no serious physical illness; (e) no history
of severe impulsivity; (f) no history of suicide attempt;
(g) not using antipsychotic medication that could affect
speech; (h) no family history of psychiatric disorder or
serious neurological disorder; and (i) provided written
informed consent to participate in the study.
2.2 Measures
We explored a smartphone APP which could record
the participant’s outgoing speech (i.e., no incoming
speech is captured or recorded). Each participant was
provided with a preloaded Samsung GALAXY Mega
6.3 (a sampling frequency of 44 kHz and a resolution
of 32 bit). Participants (both patients and controls) sat
comfortably in a noise-controlled room (background
sound below 30 dB) and were asked to use the specially
designed smartphone to call a psychiatrist from the
study and speak naturally for 15 minutes about any
topic of interest.
All call samples were saved in the Advanced Audio
Coding (aac) formant. After pre-processing of the
data, speech features of interest were extracted and
analyzed at the School of Electronic Information and
Electrical Engineering of Shanghai Jiao Tong University.
We extracted speech features with high emotional
identification rates that were identified in previous
reports of speech models of bipolar disorder [13,14]
and from our own speech model for schizophrenia
generated from data in the current study. These
parameters include prosodic features (formant 1 to 6 [F1
to F6, unit: Hz], formant bandwidth [unit: Hz], formant
amplitude [unit: dB]), and two spectral features (the
linear prediction coding [LPC], and the Mel-frequency
cepstral coefficient [MFCC]). To assess the stability of
the speech data extracted by the software, participants
in the patient group repeated the phone call 1 week
after the baseline assessment.
In addition, trained attending psychiatrists
administered the Positive and Negative Syndrome Scale
(PANSS)[15], the Scale for the Assessment of Negative
Symptoms (SANS),[16] and the Clinical Global ImpressionSchizophrenia Scale (CGI-S) [17] to the participating
patients at baseline and 1 week after baseline.
2.3 Statistical analysis
All data were processed and analyzed using SPSS 17.0
software. The in-group comparisons from baseline to
1-week post-baseline were analyzed by paired t-tests.
The test-retest reliability of the acoustic parameters
was assessed by comparing the baseline and 1-week
results using intraclass correlation coefficients (ICCs).
For between-group comparisons, continuous data with
normal distributions were analyzed using independent
t-tests; non-normal continuous variables were analyzed
using Mann-Whitney U tests; and nominal data were
analyzed using Chi-square tests. In the patient group, we
use Pearson correlation analyses to assess the strength
of the relationship between the acoustic parameters
and the severity of clinical symptoms. All statistical
analyses used two-tailed tests and statistical significance
was set at p<0.05.
3. Results
As shown in Figure 1, 26 patients completed the two
assessments. These patients included 16 males (61.5%);
• 97 •
their mean (sd) age was 43.3 (10.9) years; their mean
years of education was 9.5 (3.0) years; the mean course
of their illness was 21.7 (8.5) years; the mean number
of psychiatric admissions was 3.4 (2.4) admissions; and
the mean total length of hospitalization was 7.0 (5.4)
years. A total of 30 control subjects completed the
phonetic assessment; they included 16 males (53.3%),
had a mean (sd) age of 37.0 (14.3) years and had
a mean duration of education of 11.6 (2.5) years.
Comparison between the 26 patients who completed
the assessment and the 30 controls who completed the
assessment found no statistically significant differences
by gender (X2=0.38, p=0.536), by age (t=1.70, p=0.098),
or by duration of education (t=1.95, p=0.058).
As shown in Table 1, in the patient group there
were no statistically significant differences between
the baseline and 1-week assessment of CGI-S, PANSS
total and subscale scores, and SANS total and subscale
scores. Thus the patients’ clinical status was stable over
the 1-week interval.
Table 2 shows the baseline and 1-week results for
the acoustic parameters in the patient group and the
baseline acoustic parameters in the control group. The
test-retest reliability of these measures (only assessed
in the patient group) was good, with ICC values ranging
from 0.55 to 0.88. The prosody features and formant
amplitude and bandwidth were not significantly
different between patients and controls at baseline,
but the two spectral features were different between
the groups: MFCC was significantly lower in the patient
group than in the control group and the LPC was
significantly higher in the patient group than in the
control group.
Table 3 shows the correlation of 17 clinical and
demographic measures with the 10 acoustic parameters
in the 26 patients. Among the 170 associations
considered, ten coefficients were >0.40 and, thus,
statistically significant at the p<0.05 level: Formant
1 was negatively correlated with the SANS alogia
subscale score, Format 2 was negatively correlated
with the PANSS negative symptoms subscale score
and the SANS alogia subscale score; Formant 6 was
significantly more prominent in male than female
respondents; bandwidth was negatively correlated with
the SANS affective blunting subscale score and stronger
in female respondents than in male respondents;
and MFCC was positively correlated with the PANSS
general psychopathology subscale score and with the
patients’ total number of hospitalizations, and it was
more prominent in male respondents than female
respondents.
4. Discussion
4.1 Main findings
We found that when the severity of psychiatric
symptoms remains stable, the speech features selected
to assess the emotional content of the voice samples
of patients with schizophrenia with prominent negative
• 98 •
35 inpatients with negative symptoms of
schizophrenia treated at the Shanghai Mental
Health Center from September 2013 to
December 2015
35 healthy volunteers recruited from the
community by advertisement from September
2013 to December 2015
1 refused to participate
34 inpatients were selected for the study
35 healthy volunteers selected as the study
controls
6 were excluded:
• 2 had psychoactive
substance abuse
• 2 had bipolar disorder
• 1 had brain disease
• 1 had intellectual disability
5 were excluded:
• 2 had family history of
psychosis
• 1 had brain disease
• 1 had psychoactive
substance abuse
• 1 had personality disorder
28 inpatients were enrolled in the study
30 controls were enrolled in the study
2 failed to meet phonetic extraction
time requirements
26 inpatients completed the phonetic assessment
and clinical evaluations with the PANSS, SANS,
and CGI-S
30 controls completed the phonetic assessment
PANSS, Positive and Negative Syndrome Scale[15]
SANS, Scale for the Assessment of Negative Symptoms[16]
CGI-S, Clinical Global Impression-Schizophrenia scale[17]
Table 1. Comparisons of clinical symptoms in 26 patients with schizophrenia at baseline and after 1 week
CGI-S
PANSS
total score
positive symptoms score
negative symptoms score
general psychopathology score
SANS
total score
affective blunting score
alogia score
avolition score
anhedonia score
attention score
baseline
mean (sd)
after 1 week
mean (sd)
paired
t-test
p-value
4.69 (0.74)
--69.23 (9.62)
8.15 (1.54)
28.23 (4.03)
32.88 (5.83)
--76.00 (7.43)
26.27 (3.77)
14.15 (1.83)
15.92 (2.11)
19.27 (3.62)
0.42 (1.03)
4.73 (0.78)
--68.96 (9.64)
8.23 (1.53)
27.81 (4.24)
32.96 (5.86)
--74.92 (9.83)
24.38 (4.04)
14.04 (2.36)
16.42 (2.63)
19.73 (3.53)
0.38 (0.85)
1.00
--1.16
1.44
2.03
1.00
--0.87
1.94
0.37
1.64
1.95
1.00
0.327
--0.258
0.161
0.054
0.327
--0.391
0.064
0.713
0.114
0.063
0.327
• 99 •
Table 2. Comparisons of speech features at baseline and after 1 week in the patient group and between
patient and control groups at baseline
phonetic parameter
F1 (Hz/dB)
F2 (Hz/dB)
F3 (Hz/dB)
F4 (Hz/dB)
F5 (Hz/dB)
F6 (Hz/dB)
formant bandwidth (Hz)
formant amplitude (dB)
MFCC
LPC
baseline
patient group
result
(n=26)
mean (sd)
0.036 (0.007)
0.040 (0.052)
0.174 (0.015)
0.265 (0.023)
0.359 (0.020)
0.426 (0.011)
18.83 (11.05)
0.043 (0.005)
-0.085 (0.500)
0.249 (0.067)
patient group
test-retest
result after
reliability of
1 week
patient results
(n=26)
ICC (p-value)
mean (sd)
0.237 (0.028)
0.76 (0.002)
0.084 (0.009)
0.68 (0.009)
0.179 (0.014) 0.84 (<0.001)
0.264 (0.019)
0.67 (0.011)
0.359 (0.014) 0.88 (<0.001)
0.426 (0.012)
0.55 (0.045)
21.26 (12.41)
0.63 (0.019)
0.042 (0.004)
0.61 (0.024)
-0.027 (0.193) 0.87 (<0.001)
0.035 (0.006)
0.72 (0.004)
baseline
control group
result
(n=30)
mean (sd)
0.040 (0.005)
0.083 (0.009)
0.182 (0.012)
0.257 (0.019)
0.357 (0.013)
0.426 (0.015)
18.10 (9.81)
0.041 (0.009)
0.236 (0.043)
-0.203 (0.367)
comparison of
baseline patient v.
control group
results
t (p-value)
1.78 (0.081)
0.95 (0.353)
1.43 (0.153)
0.35 (0.725)
0.21 (0.836)
0.06 (0.951)
1.16a (0.247)
1.01 (0.317)
4.97 (<0.001)
5.69 (<0.001)
ICC, Intraclass Correlation Coefficient MFCC, Mel-frequency cepstral coefficient
LPC, linear prediction coding
a
The homogeneity of variances tests showed that the data were heterogeneous, so comparison of the results of the two groups used
the Mann Whitney U test; this is the Z-value of the Mann-Whitney U
Table 3. Correlation of demographic characteristics and the severity of clinical symptoms (at baseline) with
the speech features in 26 patients with schizophrenia (Pearson's r)
CGI-S
PANSS
total score
positive symptoms score
negative symptoms score
general psychopathology
score
SANS
total score
affective blunting score
alogia score
avolition score
anhedonia score
attention score
Age
Gender (1=female, 2=male)
Years of education
Duration of illness
Number of hospitalizations
Total time hospitalized
F1
-0.28
F2
-0.20
F3
0.04
F4
-0.10
-0.30 -0.38
0.11 0.04
-0.23 -0.40a
-0.16
0.09
-0.21
-0.26
0.18
-0.30
-0.003 0.15
0.23 -0.12
0.05 0.10
-0.14
-0.23
-0.01
0.08
-0.25
0.17
0.31
-0.03
0.13
0.09
-0.36
0.26
-0.36
-0.36
-0.15
-0.28
-0.11
0.21
-0.16
0.08
0.43a
0.06
-0.09
-0.04
-0.42a
-0.04
0.09
-0.04
-0.05
-0.18
-0.18
0.07
-0.23
-0.04
0.07
0.28
-0.42a
0.12
0.10
-0.34
0.22
0.18
-0.22
0.30
-0.13
0.10
-0.04
-0.30
0.17
0.05
0.09
0.22
0.28
-0.08
-0.10
0.16
0.10
0.10
-0.18
-0.50a
-0.01
0.08
0.11
0.08
0.05
0.03
-0.003
-0.01
-0.16
0.18
0.04
0.07
0.10
-0.02
0.02
-0.22
0.17
-0.19
0.179
-0.04
-0.01
-0.11
0.02
0.31
0.23
-0.17
-0.24
-0.23
-0.27
0.45a
-0.25
-0.25
-0.06
-0.08
-0.20
-0.41a
0.06
-0.04
-0.06
-0.23
-0.09
-0.50b
0.17
-0.25
-0.26
-0.13
0.002
-0.001
0.26
0.08
-0.18
0.06
-0.18
0.20
0.11
-0.21
-0.27
0.08
0.06
0.15
0.20
0.02
-0.03
-0.24
-0.04
0.69b
0.07
-0.01
0.40a
0.22
0.02
-0.12
-0.11
0.09
0.12
0.18
-0.05
-0.08
0.04
0.01
-0.06
0.09
MFCC, Mel-frequency cepstral coefficient
LPC, linear prediction coding
a
b
F5
0.03
0.01<p<0.05
0.001<p<0.01
F6
-0.13
bandwidth amplitude MFCC LPC
-0.13
0.19
0.04 -0.03
• 100 •
symptoms were also stable over a 1-week period.
Correlation analyses of these measures with clinical
and demographic characteristics of the patients
identified several potentially important relationships, a
finding that has been reported in previous studies.[8-10]
Comparison of these speech features between patients
and matched healthy controls found no statistically
significant differences in the prosody features or
formant bandwidth and amplitude, but there were
significant differences in the two spectral features
considered: the MFCC was significantly lower in patients
than controls, while the LPC was significantly higher in
patients than controls.
Other studies have reported that these two
spectral features play an important role in everyday
communications.[18] Spectral features have also been
found to be useful for discriminating emotions in
artificial intelligence studies. The LPC is a relatively
efficient and accurate measure of the waveform and
spectrum of speech that is used in speech coding,
speech synthesis, speech identification, and other
applications.[19,20] The MFCC, which modifies external
signals in a manner similar to the human ear, is a reliable
parameter for discriminating different emotional
states.[21,22] Similar to our results, a study by Sun and
colleagues[23] found that (when using a sorter based on
a Gaussian mixture model [GMM] algorithm) the MFCC
was better at discriminating different emotional states
than the prosody features. Further work is needed to
determine whether or not these spectral features can be
used as biomarkers for the identification and monitoring
of schizophrenia or of the prominent negative symptom
subtype of schizophrenia.
The correlation analysis identified some intriguing
associations between 6 of the 10 speech signal features
considered (F1, F2, F4, F6, bandwidth, and MFCC)
and 6 of the 17 clinical and demographic parameters
considered (gender, the PANSS negative symptoms
and general psychopathology subscale scores, the
SANS affective blunting and alogia subscale scores,
and the number of hospitalizations). Other studies
have also identified significant correlations between
different speech features and the negative symptoms
of schizophrenia.[11] This raises the possibility that a
subset of acoustic parameters of a standardized speech
sample – potentially transmitted over a smart phone to
clinicians – could be used to either monitor the severity
of negative symptoms or predict the subsequent course
of the illness. However, given the small sample size and
the large number of potential associations considered
in the current study, these results need to be replicated
before they can be meaningfully interpreted.
4.2 Limitations
The present study has several limitations that need
to be considered when interpreting the results. The
speech features selected may not be the most sensitive
measures of changes in chronic schizophrenia; further
research using a wider range of measures will be
needed to find other, potentially more sensitive,
measures. The 1-week interval we used to assess the
test-retest reliability of the speech features indicated
short-term stability in the phonetic parameters, but
we are uncertain how stable such measures are over a
longer time period. A total of 170 correlations between
the 10 speech features assessed and 17 clinical and
demographic characteristics are considered, so the
statistically significant relationships identified may be
chance findings; repeat studies are needed to confirm
their importance. The comparison between patients
and controls was cross-sectional so we cannot report
on the sensitivity of the speech features to changes in
clinical symptoms; longitudinal studies that compare
changes in the speech features to changes in the clinical
measures will be needed to determine their potential
utility as biomarkers of clinical changes. All the patients
included in the study had a prolonged course of illness
and had been on antipsychotic medication for many
years, so it is possible that this may have had an effect
on the assessed speech features. [15,16,18] Finally, the
sample was quite small – only 26 patients – so some of
the negative findings (e.g., failure to identify differences
between patients and controls) may have been due to
Type II errors.
4.3 Importance
This study focused on the negative symptoms of
schizophrenia, symptoms that are often not improved
with standard antipsychotic medications and that often
predict a poor prognosis and progressive deterioration
in social functioning. [15] The study is a preliminary
assessment of the feasibility of using speech features
that assess the emotional characteristics of speech as
biomarkers for the severity of negative symptoms in
schizophrenia and, thus, as potential predictors of the
prognosis of the disorder. The selected speech features
included both the prosodic variables used in prior
studies (i.e., rate, volume, rhythm, etc.) and two spectral
features (MFCC and LPC) that have previously been
shown to be useful in the emotional characterization
of speech samples. These speech features proved to be
stable (over a short period), some of them – the spectral
features rather than the prosody features reported in
some previous studies – were significantly different
between patients and controls, and some of them were
significantly correlated with clinical measures of negative
symptoms. However, this was a cross-sectional study in a
small group of chronic patients, so much more detailed
studies using larger samples of more diverse patients
that are followed over time will be needed before the
potential utility of such speech features can be fully
assessed.
Funding
Shanghai public health outstanding academic leader
training program (GWDTR201230); Shanghai Jiao Tong
University Key Program for the Medical Engineering
Cross Project (YG2012ZD04);and Shanghai Key
Laboratory of Psychotic Disorders (13dz2260500).
The authors report no conflict of interest.
Informed consent
Every participant who participated in this study signed a
consent form at the beginning of the study.
• 101 •
Ethical review approval
The study has been approved by the Shanghai Mental
Health Center Institutional Review Board (approval
number: 2011-15).
Authors' contributions
DHC designed the study and revised the manuscript;
J Zhang and ZP recruited patients and healthy
controls, gathered speech data, and conducted clinical
evaluations; J Zhu and CG conducted the phonetic
analyses; J Zhang did the statistical analyses and wrote
the first draft of the manuscript.
精神分裂症患者语音信号的临床分析
张静 , 潘忠德 , 桂超 , 朱杰 , 崔东红
背景：语音参数是精神分裂症生物学指标研究的一个
全新领域，其中一些已被证明与情感反应相关，情感
反应是精神分裂症患者显著受限制的一个特点，特别
是对那些具有突出阴性症状的患者。
目标：评估慢性精神分裂症患者的选择性语音参数与
临床症状严重程度之间的关系，并比较患者与所匹配
的健康对照者的这些特征。
方法：对 26 例住院慢性精神分裂症患者（入组时和
一周后）和 30 名健康对照者（仅在入组时）通过电话
采集的 15 分钟语音样本，对该样本进行 10 项语音测
量参数的评估，包括 6 个语音韵律参数、共振峰带宽
和振幅、以及 2 个频谱特征。采用阳性与阴性症状量
表 (Positive and Negative Syndrome Scale)、阴性症状评
估量表 (Scale for the Assessment of Negative Symptoms)
、临床总体印象量表 - 精神分裂症分量表 (the Clinical
Global Impression-Schizophrenia scale) 分别在基线和 1 周
后进行患者临床特征的评估。
结果：患者组症状在 1 周的时间间隔中保持稳定，并
且 10 项语音参数的前后一周重测信度良好（内部相
关系数 [intraclass correlation coefficient, ICC] 介于 0.55
到 0.88 之间）。语音参数中 6 项韵律参数、共振峰带
宽和振幅参数在患者组和对照组之间没有显著差异，
但 2 项光谱参数在组间有差异：患者组美尔频率倒谱
系数 (the Mel-frequency cepstral coefficient, MFCC) 评分
显著低于对照组，并且患者组的线性预测系数 (linear
prediction coding, LPC) 评分显著高于对照组。在患者组
中，在 10 个本研究所考虑的语音参数和 17 个所考虑
的临床参数之间构成的 170 个相关性中，有 10 个达到
了 p<0.05 的统计学显着性水平（相关系数 >0.40)。
结论：这项研究支持了语音参数具有作为精神分裂症
阴性症状严重程度指标（即，生物指标）的潜在价值
，但在这些语音参数的潜在效用获充分评估前，我们
需要对更多样化的患者进行更大样本量、更详细的随
访研究。
关键词：精神分裂症；语音；生物标志；阴性症状；
中国
本文全文中文版从 2016 年 8 月 25 日起在 http://dx.doi.org/10.11919/
j.issn.1002-0829.216025 可供免费阅览下载
References
1.
Green MF, Bearden CE, Cannon TD, Alan PF, Hellemann GS,
Horan WP, et al. Social cognition in schizophrenia, part 1：
performance across phase of illness. Schizophr Bull. 2012;
38(4): 854-864. doi: http://dx.doi.org/10.1093/schbul/
sbq171
2.
American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders, Fifth Edition (DSM-5). Arlington,
3.
Zhao L, Jiang CH, Zou CR, Wu ZY. [A study on emotional
feature analysis and recognition in speech signal]. Dian
Zi Xue Bao. 2004; 32(4): 606-608. Chinese. doi: http://
dx.chinadoi.cn/10.3321/j.issn:1000-436X.2000.10.004
4.
Nwe TL, Foo SW, Silva LCD. Speech emotion recognition
using hidden Markov models. Speech Communication.
2003; 41(3): 603-623. doi: http://dx.doi.org/10.1016/S01676393(03)00099-2
5.
Lin YL, Wei G, Yang KC. [A survey of emotion recognition
in speech]. Guangzhou Dian Lu Yu Xi Tong Xue Bao. 2007;
12(1): 90-91. Chinese. doi: http://dx.chinadoi.cn/10.3969/
j.issn.1007-0249.2007.01.019
6.
Yuan J, Xu HH, He X. [Research progress of speech emotion
recognition]. Ji Suan Ji Guang Pan Ruan Jian Yu Ying Yong.
2010; 1: 36-38. Chinese
• 102 •
7.
Bhatti MW, Wang Y, Guan L. A neural network approach
for human emotion recognition in speech (conference
paper). Sydney: Circuits and Systems Conference, ISCAS
‘04; 2004. p.181-184. doi: http://dx.doi.org/10.1109/
ISCAS.2004.1329238
8.
Stassen HH, Albers M, Püschel J, Scharfetter C, Tewesmeier
M, Woggon B. Speaking behavior and voice sound
characteristics associated with negative schizophrenia.
J Psychiatr Res. 1995; 29(4): 277-296. doi: http://dx.doi.
org/10.1016/0022-3956(95)00004-O
9.
Püschel J, Stassen HH, Bomben G, Scharfetter C, Hell D.
Speaking behavior and speech sound characteristics in
acute schizophrenia. J Psychiatr Res. 1998; 32(2): 89-97. doi:
http://dx.doi.org/10.1016/S0022-3956(98)00046-6
10. Leitman DI, Laukka P, Juslin PN, Saccente E, Butler P, Javitt
DC. Getting the cue: sensory contributions to auditory
emotion recognition impairments in schizophrenia.
Schizophr Bull. 2008; 36(3): 545-556. doi: http://dx.doi.
org/10.1093/schbul/sbn115
11. Gold R, Butler P, Revheim N, Leitman DI, Hansen JA, Gur
RC, et al. Auditory emotion recognition impairments
in schizophrenia: relationship to acoustic features and
cognition. Am J Psychiatry. 2012; 169(4): 424-432. doi:
http://dx.doi.org/10.1176/appi.ajp.2011.11081230
12. Sheehan DV, Lecrubier Y, Sheehan KH, Amorim P, Janavs
J, Weiller E, et al. The Mini-International Neuropsychiatric
Interview (M.I.N.I.): the development and validation of a
structured diagnostic psychiatric interview for DSM-IV and
ICD-10. J Clin Psychiatry. 1998; 59 (Suppl 20): 22-33
13. Xu D, Zhang J, Zhu J, Cui D. Acoustic Analysis and
Identification of Manic Psychosis Patients. London:
Intelligent Signal Processing Conference; 2013
14. Gui C, Li W, Pan Z, Zhang J, Zhu J, Cui D. A classifier for
diagnosis of manic psychosis state based on SVM-GMM.
Sydney: The 10th International Conference on Information
Technology and Applications (ICITA2015); 2015
15. Kay SR, Fiszbein A, Opler LA. The positive and negative
syndrome scale (PANSS) for schizophrenia. Schizophr Bull.
1987; 13(2): 261-276. PMID 3616518. doi: http://dx.doi.
org/10.1093/schbul/13.2.261.
16. Andreasen NC. Negative symptoms in schizophrenia:
definition and reliability. Arch Gen Psychiatry. 1982; 39(7):
784-788
17. Haro J, Kamath S, Ochoa S, Novick D, Rele K, Fargas A, et al.
The Clinical Global Impression-Schizophrenia scale: a simple
instrument to measure the diversity of symptoms present
in schizophrenia. Acta Psychiatrica Scandinavica. 2003;
107(Suppl 416): 16-23. doi: http://dx.doi.org/10.1034/
j.1600-0447.107.s416.5.x
18. Sun Y, Jiang ZC, Wang DF. [Analysis and application of
voice spectrum]. Ji Suan Ji Yu Xian Dai Hua. 2010; 4:
j.issn.1006-2475.2010.04.054
19. Yu BK, Yu M. [Analysis of the resonance peak of speech
signal extracted with LPC method]. Dian Sheng Ji Shu.
2000; 3: 3-8. Chinese. doi: http://dx.chinadoi.cn/10.3969/
j.issn.1002-8684.2000.03.001
20. Nica A, Caruntu A, Toderean G, Buza O. Analysis and
synthesis of vowels using Matlab. IEEE Conference on
Automation, Quality and Testing, Robotics; 2006. p. 371-374.
doi: http://dx.doi.org/10.1109/AQTR.2006.254662
21. Lin W, Yang LL, Xu BL. [Speaker identification in Chinese
whispered speech based on modified-MFCC]. Nanjing
Da Xue Xue Bao (Zi Ran Ke Xue). 2006; 42(1): 5461. Chinese. doi: http://dx.chinadoi.cn/10.3321/
j.issn:0469-5097.2006.01.008
22. Dave N. Feature extraction methods LPC, PLP and MFCC in
speech recognition. Ijaret Org. 2013; 1(6): 1-5
23. Sun MH, Jiang BC. [Analysis and Research on the Emotional
Information of Mandarin Speech]. Shandong: Shandong
University. 2011; p. 22-24. Chinese
Dr. Jing Zhang graduated from Wannan Medical College with a bachelor’s degree in medicine in 2007.
She is now a master’s student at the Shanghai Mental Health Center, Shanghai Jiao Tong University
School of Medicine. She has been working in the Department of Psychiatry as an attending physician
in the Jiading District Mental Health Center in Shanghai since 2009. Her main research interest is the
diagnostic relevance of speech signals in patients with schizophrenia.
• 103 •
•Forum•
Is the DSM-5 hoarding disorder diagnosis valid in China?
Zhen WANG, Yuan WANG, Qing ZHAO, Kaida JIANG*
Summary: Hoarding disorder, newly included as a separate diagnostic entity in the Obsessive-Compulsive
and Related Disorders section of DSM-5, has been reported to have significantly different symptoms and
etiology than obsessive-compulsive disorder (OCD). However, the validity of this new diagnosis in China
– where the storing of possessions is sanctioned and normalized – remains to be proven. We considered
available data about pathological hoarding in East Asia and found the condition to be relatively common
and symptomatically similar to that reported in western countries. We conclude that the ‘Hoarding
Disorder’ diagnosis defined in DSM-5 is a valid clinical entity in China, though when making the diagnosis
clinicians must take care to differentiate pathological hoarding that is distressing to the individual and
significantly interferes with social and occupational functioning from culturally sanctioned thriftiness that is
not associated with either distress or social dysfunction.
Keywords: hoarding disorder; DSM-5; cross-cultural validity; case report; China
Hoarding behavior has long been considered one
of the symptoms of obsessive compulsive disorder
(OCD). However, recent research reporting significant
differences among individuals with pathological
hoarding, patients with OCD, and healthy controls in
symptomatology, cognitive functioning, family history,
and neuro-imaging [1,2] has prompted the American
Psychiatric Association to make hoarding disorder a
distinct condition in the recently published Fifth edition
of the Diagnostic and Statistical Manual of Mental
Disorders (DSM-5). [3] Listed as one of the separate
disorders under the new DSM-5 diagnostic group of
‘Obsessive-Compulsive and Related Disorders’, hoarding
disorder has three core symptoms: (a) persistent
difficulty discarding possessions regardless of value; (b)
the accumulation of possessions congests one’s active
living space; and (c) hoarding causes clinically significant
distress or functional impairment.
Using these criteria, estimates of the prevalence
of hoarding disorder in the general population range
from 1.4% to 5.8%.[4,5] About 40% of patients who meet
diagnostic criteria for OCD have hoarding symptoms
(though in most cases it is the not the main OCD
symptom), but 80% of individuals with pathological
hoarding do not meet the diagnostic criteria of OCD.[6,7]
In support of this decision to distinguish hoarding
disorder from OCD, a meta-analysis[8] found that routine
treatment for OCD among OCD patients with hoarding
symptoms is significantly less effective than for OCD
patients without hoarding symptoms.
However, there is still controversy about whether
or not hoarding disorder should be considered an
independent diagnosis, particularly in non-western
cultures where the storing of possessions, including
possessions of little current utility, is sanctioned and
normalized. In these settings, direct application of
the DSM-5 criteria could lead to over-diagnosis – the
medicalization of a culturally acceptable behavior. Most
of the research about hoarding has been conducted in
high-income countries in Europe and North America,
so research in non-western countries and in low- and
middle-income countries is needed to assess the crossnational and cross-cultural validity of the new diagnostic
criteria for hoarding disorder.
In Japan Matsunage and colleagues[9] reported that
among 168 patients with OCD, 54 (32%) had hoarding
symptoms; consistent with findings from outside of
correspondence: Professor Kaida Jiang, Shanghai Mental Health Center, 600 Wan Ping Nan Road, Shanghai 200030, China. E-mail: [email protected]
• 104 •
Asia,[10] they found that compared to OCD patients
without hoarding those with hoarding had an earlier age
of onset, more serious OCD symptoms, poorer insight,
and a higher prevalence of other comorbid mental
disorders. Chasson and colleagues [11] assessed the
psychometric properties of the Mandarin version of the
Obsessive-Compulsive Inventory-Revised (OCI-R) among
Chinese OCD patients and found that the internal
consistency, test-retest reliability, and criteria validity
were all satisfactory and similar to results from other
cultural backgrounds.[12,13] In our own recent (as yet
unpublished) work, we administered the Chinese version
of the Saving Inventory-Revised (SI-R)[14] to 341 healthy
volunteers and 140 individuals receiving treatment for
a variety of mental disorders and found that hoarding
was most common in individuals with OCD and, to a
somewhat less extent, in individuals with Generalized
Anxiety Disorder (GAD). Taken together, these findings
suggest, but do not prove, that pathological hoarding is
common in East Asia and that the clinical characteristics
of the condition are similar to those reported in western
countries.
There are, however, some differences between
western and Asian results. Factor analysis of the results
of a study by Tang and colleagues[15] that administered
the Chinese SI-R scale to 2100 Chinese university
students only identified two independent factors –
‘acquisition/difficulty discarding’ and ‘clutter’; this is
different from the three factors identified in Western
samples [13] (‘acquisition’, ‘difficulty discarding’, and
‘clutter’). Tang and colleagues[15] posit that the reason
for the difference may be that in Chinese culture
‘acquisition’ and ‘not discarding’ are active and passive
aspects of the same traditional cultural concept of ‘to
save is to earn’. Timpano and colleagues[16] compared
hoarding behaviors using OCI-R and beliefs about
hoarding using a novel hoarding beliefs questionnaire
between 303 Chinese and 87 American undergraduates:
they found that the mean (sd) overall hoarding score
was significantly higher in Chinese students (25.3
[10.7]) than in American students (15.6 [11.6]). They
also reported that hoarding behaviors among Chinese
students were mainly related to two beliefs (‘it could be
useful one day’ [usefulness], and ‘nothing is supposed
to be wasted’ [wastefulness]), while the American
students had a wider range of hoarding behaviors and
beliefs (including ‘stuff could bring visual joy’ [aesthetic
qualities], ‘stuff can help to invoke specific memories’
[remembrance], and ‘one has a responsibility to keep
stuff in good condition’ [responsibility]). Our own
(unpublished) work also found relatively high levels of
self-reported hoarding behavior in healthy community
volunteers. These results suggest that there may need
to be some cultural adaptation when applying westernbased diagnostic criteria for hoarding disorder in
Asian samples and that the cutoff scores for classifying
pathological levels of hoarding when using translated
versions of western scales of hoarding behavior may
need to be revised.
‘Making the best use of everything’ and ‘avoiding
waste’ are core values in Chinese culture that emerged
in times of scarcity when preserving everything that
may potentially be of use in the future was a reasonable
strategy to enhance personal-security.[17,18] The very
high saving rates of personal and family income in
China show that these beliefs about personal and
family security have persisted despite recent dramatic
improvements in living standards. We conclude the
‘Hoarding Disorder’ is relevant in China, but care needs
to be taken to differentiate pathological hoarding that is
distressing to the individual and significantly interferes
with social and occupational functioning from culturally
sanctioned thriftiness that is not associated with either
distress or social dysfunction.
Funding
None.
The authors declared no conflict of interest related to
this manuscript.
DSM-5 囤积障碍诊断在中国是否适用？
王振，王渊，赵青，江开达
概述：囤积障碍 (hoarding disorder)，作为新近被纳入
DSM-5 强迫症和相关障碍部分的一个独立疾病，与强
迫症 (obsessive-compulsive disorder, OCD) 相比具有明显
不同的症状和病因。然而，在中国，人们认可储藏个
人财物并认为这是正常的，这种新的诊断方法在中国
的效度还有待证明。我们研究了东亚地区有关病理性
囤积的可用数据，并发现囤积是比较常见的情况，而
且出现的症状也类似于西方国家的报道。我们认为，
DSM-5 中定义的“囤积障碍”在中国是一种合理的临
床实体，虽然临床医生在作出该诊断时必须小心区分
病理性囤积与文化上所认可的节俭，前者令患者非常
痛苦并且明显妨碍其社会和职业功能，而后者与痛苦
或社交障碍都不相关的。
关键词：囤积障碍; DSM-5; 跨文化有效性；病例报告；
中国
• 105 •
References
1.
Steketee G, Frost R, Kyrios M. Cognitive aspects of
compulsive hoarding. Cogn Ther Res. 2003; 27(4): 463-479.
doi: http://dx.doi.org/10.1023/A:1025428631552
2.
Mataix-Cols D, Frost RO, Pertusa A, Clark LA, Saxena S,
Leckman JF, et al. Hoarding disorder: a new diagnosis for
DSM-V? Depress Anxiety. 2010; 27(6): 556-572. doi: http://
dx.doi.org/10.1002/da.20693
3.
American Psychiatric Association. Diagnostic and Statistical
Manual of Mental Disorders, Fifth Edition (DSM-5). Arlington,
VA: American Psychiatric Publishing; 2013
4.
Nordsletten AE, Reichenberg A, Hatch SL, Fernández de la
Cruz L, Pertusa A, Hotopf M, et al. Epidemiology of hoarding
disorder. Br J Psychiatry. 2013; 203(6): 445-452. doi: http://
dx.doi.org/10.1192/bjp.bp.113.130195
5.
Timpano KR, Exner C, Glaesmer H, Rief W, Keshaviah A,
Brahler E, et al. The epidemiology of the proposed DSM-5
hoarding disorder: exploration of the acquisition specifier,
associated features, and distress. J Clin Psychiatry. 2011; 72:
780-786. doi: http://dx.doi.org/10.4088/JCP.10m06380
6.
Van Ameringen M, Patterson B, Simpson W. DSM5 obsessive-compulsive and related disorders: clinical
implications of new criteria. Depress Anxiety. 2014; 31(6):
487-493. doi: http://dx.doi.org/10.1002/da.22259
7.
Mataix-Cols D, Frost RO, Pertusa A, Clark LA, Saxena
S, Leckman JF, et al. Hoarding disorder: a new diagnosis for
DSM-V? Depress Anxiety. 2010; 27(6): 556-572. doi: http://
dx.doi.org/10.1002/da.20693
8.
Bloch MH, Bartley CA, Zipperer L, Jakubovski E, LanderosWeisenberger A, Pittenger C, et al. Meta-analysis: hoarding
symptoms associated with poor treatment outcome in
obsessive-compulsive disorder. Mol Psychiatry. 2014; 19(9):
1025-1030. doi: http://dx.doi.org/10.1038/mp.2014.50
9.
Matsunaga H, Hayashida K, Kiriike N, Nagata T, Stein DJ.
Clinical features and treatment characteristics of compulsive
hoarding in Japanese patients with obsessive-compulsive
disorder. CNS Spectr. 2010; 15(04): 258-266
10. Torres A R, Fontenelle L F, Ferrão Y A, do Rosário
MC, Torresan RC, Miguel EC, et al. Clinical features of
obsessive-compulsive disorder with hoarding symptoms: a
multicenter study. J Psychiatr Res. 2012; 46(6): 724-732. doi:
http://dx.doi.org/10.1016/j.jpsychires.2012.03.005
11. Chasson GS, Tang S, Gray B, Sun H, Wang J. Further validation
of a Chinese version of the obsessive-compulsive inventoryrevised. Behav Cogn Psychother. 2013; 41(02): 249-254. doi:
http://dx.doi.org/10.1017/S1352465812000379
12. Sica C, Ghisi M, Altoè G, Chiri LR, Franceschini S, Coradeschi
D, et al. The Italian version of the Obsessive Compulsive
Inventory: Its psychometric properties on community and
clinical samples. J Anxiety Disord. 2009; 23(2): 204-211. doi:
http://dx.doi.org/10.1016/j.janxdis.2008.07.001
13. Huppert J D, Walther M R, Hajcak G, Yadin E, Foa
EB, Simpson HB, et al. The OCI-R: validation of the subscales
in a clinical sample. J Anxiety Disord. 2007; 21(3): 394-406.
doi: http://dx.doi.org/10.1016/j.janxdis.2006.05.006
14. Frost RO. Measurement of compulsive hoarding: Saving
Inventory-Revised. Behav Res Ther. 2004; 42(10):1163-1183
15. Tang T, Wang JP, Tang SQ, Zhao LN. [Psychometric properties
of the Saving Inventory-Revised in Chinese University
students sample]. Zhongguo Lin Chuang Xin Li Xue Za Zhi.
2012; 20(1): 7. Chinese
16. Timpano KR, Cek D, Fu ZF, Tang T, Wang JP, Chasson GS.
A consideration of hoarding disorder symptoms in China.
Compr Psychiatry. 2015; 57: 36-47. doi: http://dx.doi.
org/10.1016/j.comppsych.2014.11.006
17. Alcon J, Glazier K, Rodriguez C. From clutter to modern art:
a Chinese artist’s perspective on hoarding behaviors. Am J
Psychiatry. 2011; 168(12). doi: http://dx.doi.org/10.1176/
appi.ajp.2011.11091414
18. King AYC. The individual and group in Confucianism: a
relational perspective. In: & Munro DJ, editor. Individualism
and holism: Studies in Confucian and Taoist Values. Ann
Arbor: Centre for Chinese Students, University of Michigan;
1985. p. 57-70
Dr. Zhen WANG received his medical bachelor’s degree from Jining Medical School in 2000, his
medical master’s degree from Shanghai Jiao Tong University in 2003, and his PhD from Shanghai
Jiao Tong University in 2009. Since graduation he has worked as a psychiatrist in the Shanghai
Mental Health Center where he is currently an associate professor and the director of the Research
and Service Department. His main research interests are the etiology and treatment of obsessivecompulsive disorder and stress and trauma-related disorders.
• 106 •
•Case report•
Behavioral and emotional manifestations in a child with
Prader-Willi syndrome
Satyakam MOHAPATRA*, Udit Kumar PANDA
Summary: Prader-Willi syndrome is a neurodevelopmental disorder characterized by mental retardation
and distinct physical, behavioral, and psychiatric features. Maladaptive behaviours, cognitive impairment,
and impediments in speech and language seriously affect the early development and long-term functioning
of individuals affected by the illness. We present a case of a 9-year-old child with Prader-Willi syndrome
whose behavioural symptoms were treated with low-dose antipsychotic medications.
Keywords: Prader-Willi Syndrome; psychiatric symptoms; childhood disorders; case report; India
1. Introduction
Prader-Willi syndrome (PWS) is a genetically determined
neurodevelopmental disorder with a prevalence of
3 to 7 individuals per 100,000 births.[1] It is usually
the result of a paternally transmitted deletion at
chromosome 15-q11-q13. Characterized by mental
retardation and distinct physical, behavioral, and
psychiatric features, individuals with PWS are typically
short and obese, have small hands and feet, and have
other dysmorphic features including a narrow bifrontal
diameter, full cheeks, and almond-shaped eyes.[2] They
have borderline to moderate mental retardation, have
impaired speech and language, [3] and exhibit more
behavioral disturbances than individuals with other
intellectual disabilities,[4] including excessive interest
in food, skin picking, difficulty with changes in routine,
temper tantrums, obsessive and compulsive behaviors,
and mood fluctuations.[5,6] The severity of the behavioral
problems increases with age and body mass index,[7]
and then diminishes in older adults.[9] Recent evidence
suggests that autism spectrum disorders (ASD) may be
common in individuals with PWS.[9] Psychosis occurs
during young adulthood in 5-10% of individuals with the
syndrome.[10] The cognitive impairment, limited speech
and language skills, and behavioral abnormalities
seriously affect the early development and long-term
functioning of individuals with PWS. Psychiatric and
behavioral problems are the most common cause of
hospitalization.
2. Case history
A 9-year-old girl was brought to our hospital with
complaints of irritability, stubbornness, emotional
lability, temper tantrums, and increased speech. Her
father also reported hyperactivity, a history of overeating and stealing food, and sudden mood changes
including outbursts of laughter and crying without any
obvious reason. She had had an uneventful birth history
and no family history of neurological or psychiatric
illness, but she had delayed development of gross
motor functions and language skills. Her academic
performance in primary school was poor.
Physical examination revealed an obese female
(weight 54 kg, height 112 cm, body mass index 43.1)
with small hands and feet, a narrow nasal bridge, and
Mental Health Institute, Sriram Chandra Bhanj (SCB) Medical College, Cuttack, Odisha, India
*correspondence: Dr. Satyakam Mohapatra, Mental Health Institute, S.C.B. Medical College, Cuttack - 753 007, Odisha, India.
Email: [email protected]
almond-shaped palpebral fissures. The skin on her
face, hands, and arms had many excoriated papules
from repetitive skin picking. Her speech had imprecise
articulation and hypernasality. Her neurological
examination, routine blood tests, thyroid function tests,
liver function tests, and computed tomography (CT) of
the brain were all within normal limits. Ultrasonography
of her abdomen and pelvis revealed a fatty liver and a
hypoplastic uterus.
Intelligence testing indicated that her IQ was
40. After consultation with an endocrinologist, she
was diagnosed with Prader-Willi syndrome. Her
management included hormonal therapy and dietary
advice directed by the endocrinology department, skin
treatment directed by the dermatology department,
and speech therapy. She was also given risperidone
1 mg/d for behavioral control. Her family was educated
about the illness. After 8 weeks of this multi-phased
intervention, her irritability, stubbornness, temper
tantrums, increased speech, and self- injurious behavior
improved significantly. She tolerated the risperidone
well without any significant adverse reaction. After
4 months of treatment the dose of risperidone was
reduced to 0.5 mg per day.
3. Discussion
In the past the Prader-Willi syndrome was diagnosed
based on the clinical presentation, but genetic testing
can now more accurately diagnose the condition. In
high-income countries, genetic testing is recommended
for all infants with pronounced hypotonia; however, in
most low- and middle-income countries genetic testing
is not available, so the diagnosis still depends on the
correct identification of the typical clinical symptoms.
Given the relative rarity of the disorder and the
unfamiliarity of most clinicians with the condition, many
cases go undiagnosed.
It is not feasible to correct the genetic abnormality,
so most treatments are aimed at suppressing unwanted
symptoms. Given the frequent occurrence of difficult-
• 107 •
to-manage behavioral problems in PWS, clinicians
often try low-dose antipsychotic medication. One study
unexpectedly found that antipsychotic medications
– which often lead to weight gain in patients with
schizophrenia – was associated with weight loss in
patients with PWS. [11] However, the small numbers
of individuals with PWS make it difficult to conduct
formal evaluations of the effectiveness of antipsychotic
medications or other interventions, so it has not
been possible to develop evidence-based treatment
guidelines for the condition. In most cases, clinicians
must use their judgment to individualize the treatment
to the needs of each patient. As shown in the current
case, the ongoing involvement of multiple disciplines
along with educational and psychological support for
the care-givers is often needed to address the complex
needs of these patients and their families.
Funding
No funding support was obtained for preparing this case
report.
The authors declare that they have no conflict of
interest related to this manuscript.
Informed consent
The father of the patient signed an informed consent
form and agreed to the publication of this case report
SM drafted the manuscript. UKP critically reviewed the
manuscript. SM and UKP both carried out the clinical
diagnosis and the psychiatric evaluation. Both authors
read and approved the final manuscript.
Prader-Willi 综合征患儿的行为与情绪表现
Satyakam M, Panda UK
概述：Prader-Willi 综合征是一种神经发育障碍，以
精神发育迟滞以及明显的躯体、行为与精神方面的
表现为特征。适应不良性行为、认知损害以及言语
和语言障碍严重影响患者早期发育，也会影响患者
的长期功能。本文报告一例 9 岁的 Prader-Willi 综合
征患儿，以低剂量抗精神病药物治疗其行为症状。
关键词：Prader-Willi 综合征；精神症状；儿童期障碍 ;
病例报告；印度
References
1.
Cassidy SB, Driscoll DJ. Prader-Willi syndrome. Eur J Hum
Genet. 2009; 17(1): 3-13. doi: http://dx.doi.org/10.1038/
ejhg.2008.165
2.
Holm VA, Cassidy SB, Butler MG, Hanchett JM, Greenswag
LR, Whitman BY, et al. Prader-Willi syndrome: consensus
diagnostic criteria. Pediatrics. 1993; 91: 398–402
• 108 •
3.
Lewis BA, Freebairn L, Heeger S, Cassidy SB. Speech and
language skills of individuals with Prader-Willi syndrome.
Am J Speech-Language Pathol. 2002; 11: 285–294. doi:
http://dx.doi.org/10.1044/1058-0360(2002/033)
7.
Steinhausen HC, Eiholzer U, Hauffa BP, Malin Z. Behavioral
and emotional disturbances in people with Prader-Willi
syndrome. J Intellect Disabil Res. 2004; 48(1): 47–52. doi:
http://dx.doi.org/10.1111/j.1365-2788.2004.00582.x
4.
Curfs LM, Verhulst FC, Fryns JP. Behavioral and emotional
problems in youngsters with Prader-Willi syndrome. Genet
Couns. 1991; 2(1): 33–41
8.
Dykens EM. Maladaptive and compulsive behavior in
Prader-Willi syndrome: new insights from older adults. Am
J Ment Retard. 2004; 109(2): 142–153
5.
Holland AJ, Whittington JE, Butler J, Webb T, Boer H, Clarke
D. Behavioral phenotypes associated with specific genetic
disorders: evidence from population based study of people
with Prader-Willi syndrome. Psychol Med. 2003; 33(1):
141–153
9.
Veltman MW, Craig EE, Bolton PF. Autism spectrum
disorders in Prader-Willi and Angelman syndromes: a
systematic review. Psychiatr Genet. 2005; 15(4): 243–254
10.
Vogels A, Van Den Ende J, Keymolen K, Mortier G, Devriendt
K, Legius E, et al. Psychotic disorders in Prader-Willi
syndrome. Am J Med Genet A. 2004; 127A(3): 238–243.
doi: http://dx.doi.org/10.1002/ajmg.a.30004
11.
Elliott JP, Cherpes G, Kamal K, Chopra I, Harrison
C, Riedy M, et al. Relationship between antipsychotics
and weight in patients with Prader–Willi syndrome.
Pharmacotherapy. 2015; 35(3): 260-268
6.
Einfeld SL, Smith A, Durvasula S, Florio T, Tonge BJ. Behavior
and emotional disturbance in Prader-Willi syndrome. Am
J Med Genet. 1999; 82(2): 123–127. doi: http://dx.doi.
org/10.1002/(SICI)1096-8628(19990115)82:2<123::AIDAJMG4>3.0.CO;2-C
Dr. Mohapatra obtained his bachelor’s degree from MKCG Medical College, Berhampur, Odisha, India
in 2007 and his MD from King George’s Medical University, Lucknow, India in 2012. He is currently
working as a senior resident in the Department of Psychiatry in the Mental Health Institute in Sriram
Chandra Bhanj (SCB) Medical College in Odisha. His main research interests are psychopharmacology
and child psychiatry.
• 109 •
•Case report•
Treatment resistant depression or dementia: a case report
Zhongyong SHI1,2, Shifu XIAO1,3,*, Xia LI1,3,*
Summary: The co-occurrence of depression and dementia is increasingly common in the elderly.
The current case describes a 78-year-old female with two previous episodes of major depression
who presented with both symptoms of depression (amotivation and flattened affect) and typical
symptoms of dementia (impaired memory and executive functioning). Even after a detailed clinical
exam and neuropsychiatric testing, it remained difficult to definitively classify the diagnosis as either
treatment-resistant depression or old-age dementia. After 8 weeks of inpatient treatment, including
changing her reserpine-based antihypertensive medication, adjusting her antidepressants, and
providing psychotherapy, her depressive and anxiety symptoms improved, but most of her cognitive
symptoms persisted. Her symptoms did not change over 7 months of post-hospitalization follow-up.
She subsequently developed advanced breast cancer and started chemotherapy; at this point her
depressive and cognitive symptoms became more pronounced. We conclude that it will take two-tothree years of follow-up to determine whether the cognitive symptoms are residual to her depression
or a newly emerging dementia (or both). This case shows that for elderly patients who have symptoms
of both depression and dementia, detailed clinical examination and neuropsychiatric testing may need
to be combined with longitudinal assessment of their responsiveness to treatment before a definitive
diagnosis can be assigned.
Keywords: depression; dementia; pseudo-dementia; case report; China
1. Case history
A 78-year-old female with vocational school education
was admitted for a second hospitalization after a
seven-year history of depressive symptoms, a twoyear history of reduced physical activity, and a oneyear history of memory loss. Initially, in 2008 after
being frightened by the results of a urine test for a
urinary tract infection, she became depressed and
anorexic. She was diagnosed at a local general hospital
with major depression (without psychotic symptoms)
and treated with paroxetine and alprazolam for two
months until the symptoms remitted. In 2012, while
providing nursing care for her younger sister who
had lung cancer, her depressive symptoms recurred
and she made two suicide attempts by overdose
with alprazolam. This led to a 1-month psychiatric
hospitalization that included treatment with sertraline,
mirtazapine, and electroconvulsive therapy (ECT);
after discharge she was able to do housework and
grocery shopping. However, starting in June 2013,
without any obvious trigger her activity level gradually
decreased. She became uncommunicative and
was unwilling to leave her home, neither watching
TV nor doing any housework. She spent all of her
time in bed, so her 82-year-old husband (who had
hypertension and chronic bronchitis) had to assist
with her eating and routine self-care. Starting in
2014 she reported experiencing short-term memory
loss which made her forget what she had eaten for
breakfast or what she had just read in the newspaper.
She was initially treated as an outpatient, though
most of the outpatient visits were with the husband
alone because he was physically unable to regularly
transport her to the outpatient department. She
was prescribed a variety of medications including
1
Shanghai Tenth People’s Hospital, Shanghai Tong Ji University, Shanghai, China
3
Alzheimer Diagnosis and Treatment center, Shanghai Jiao Tong University School of Medicine, Shanghai, China
2
*co-corresponding authors: Professor Shifu Xiao and Dr. Xia Li. Department of Geriatric Psychiatry, Shanghai Mental Health Center, 600 South Wan Ping
Road, Shanghai 200030, China. E-mail: (Shifu Xiao) [email protected]; (Xia Li) [email protected]
• 110 •
venlafaxine, citalopram, mianserin, Deanxit (a
combined preparation of flupentixol and melitracen),
mirtazapine, amisulpride, and quetiapine. Over
18 months of outpatient treatment there was no
significant improvement, so she was readmitted to
hospital.
On admission the patient reported a history
of hypertension which had been controlled for the
previous four years by taking 1-2 tablets per day of a
combination anti-hypertensive (each tablet included
reserpine, hydrochlorothiazide, dihydralazine, and
promethazine), but she had no other significant
personal or family history of illness. Physical
examination revealed excoriated skin with infected
lesions on her limbs and face due to poor hygiene and
scratching. On mental status exam she was conscious,
but she appeared apathetic, repeatedly asking to
be allowed to go to sleep during the interview. She
was inattentive, responding passively to questions
with answers that were devoid of meaning. She was
disoriented to time and space and had apparent
memory lapses. She reported intermittent auditory
hallucinations, saying that she could hear her brother
speaking in the doorway.
Further laboratory and neuropsychological
examinations had the following results: (a) no
abnormal blood chemistry results; (b) an abnormal
electroencephalogram (EEG) (shown in Figure 1,
Panel A), with the δ domain power increased and the
α domain power decreased; (c) abnormal magnetic
resonance imaging (MRI) test (shown in Figures 2
and 3), indicating ischemia in the right basal ganglia
area and bilateral frontal lobe, hippocampus atrophy,
and diffuse conical atrophy; (d) mild depression and
anxiety as rated by the Hamilton Rating Scale
for Depression (HAMD)[1] (score=15), and the Hamilton
Anxiety Rating Scale (HAMA) [2] (score=8); and (c)
decline in cognitive function as indicated by results
of the Mini Mental State Examination (MMSE) [3]
(score=21), the Montreal Cognitive Assessment Scale
(MoCA)[4] (score=11), the Wechsler Adult Intelligence
Scale-Revised in Chinese (WASI-RC)[5] (IQ score=64),
the revised Chinese version of the Wechsler Memory
(WMSRC) [6] (score=47), and supplemental tests
that revealed difficulties in delayed memory, visual
recognition, and verbal fluency.
The admitting diagnosis was major depression
and possible dementia (which had to be ruled out).
On admission she continued on the medications she
had been taking as an outpatient: citalopram 20 mg/d,
mirtazapine 15 mg/d, quetiapine 25 mg/d, memantine
5 m g / d , ox i ra c e ta m 4 0 0 m g b i d ( to p ro m o te
recognition), and lorazepam 0.5 mg qn (to improve
sleep). Because reserpine, the anti-hypertensive
medication she had been taking for four years, is a
monoamine depletion agent which may exacerbate
depression, we changed it to amlodipine 5 mg/d to
control her blood pressure. While an inpatient she also
participated in group cognitive behavioral therapy,
family counseling, and art and music activities.
There was no significant improvement after 2 weeks
of treatment: her drowsiness, limited speech, lack
of interest, and aversion to activity persisted. So
we modified the treatment plan to the following:
sertraline 50 mg/d, memantine 10 mg/d, rivastigmine
3 mg bid, and aripiprazole 2.5 mg/d. After 6 weeks of
this new treatment regimen her test results indicated
improvement in the depressive and anxiety symptoms:
HAMD=9; HAMA=6. Her EEG had returned to normal,
as shown in Panel B of Figure 1, the fragmentary
Figure 1. Electroencephalogram (EEG). (A) Abnormal EEG on admission: prompting θ, δ domain power
increased, and α domain power decreased. (B) EEG returns to normal after 6 weeks of treatment:
θ domain power in occipitoparietal region increased and α domain power normalized
• 111 •
Figure 2. Cranial magnetic resonance imaging: right basal ganglia area and bilateral frontal area have
punctiform ischemia (A is T1 weighted image; B is T2 weighted image)
Figure 3. Hippocampus magnetic resonance imaging: hippocampus atrophy and brain atrophy (A is coronal
view, B is sagittal view)
auditory hallucinations had faded, and she had
become somewhat more socially interactive. However,
her cognitive problems with attention and executive
functioning had only improved slightly: MMSE=22; and
MoCA=16.
At the time of hospital discharge, 8 weeks after
admission, her discharge diagnosis was treatment
resistant depression and the discharge medications
were amlodipine 5 mg/d, sertraline 100 mg/d, and
rivastigmine 3 mg bid. After she returned home, we
also arranged for the delivery of daily meals and for
regular home visits from community workers. She
attended monthly outpatient follow-up visits, but
there was no further improvement in her depressive
and cognitive symptoms. Seven months after discharge
she became increasingly fatigued and experienced
many falls; this resulted in admission to the surgery
department of a general hospital where she was
diagnosed with advanced breast cancer and started
on chemotherapy. At that point her sertraline and
rivastigmine were stopped and she was changed to
trazadone 50 mg qn (to improve sleep). Her depressive
and cognitive symptoms subsequently increased.
• 112 •
2. Discussion
Prior to the current admission the patient had been
treated with several different antidepressants at an
adequate dose for at least 6 weeks, but none of these
courses of treatment were effective, so she met criteria
for treatment resistant depression. Unlike her previous
depressive episodes, the current episode included
physical weakness and apathy, transient hallucinatory
phenomena, and marked cognitive decline. Based
on these findings and the diffuse conical atrophy
and hippocampal atrophy seen on brain imaging, we
concluded that the patient also met diagnostic criteria
for Alzheimer’s disease(AD).[8] The diagnostic difficulty
was to determine whether she had two concurrent
illnesses or a single illness that included both
prominent affective and prominent cognitive features.
And if it was a single disorder, which one?
Considering her prior episodes of depression, our
initial diagnosis was depression with impaired cognitive
functioning. Epidemiological studies[9] report that 3050% of patients with depression have concurrent
cognitive dysfunction; in elderly depressed patients
the cognitive dysfunction can include impairments
in memory, attention, information processing, and
executive functioning.[10] If the cognitive dysfunction
in depressed elderly is improved with antidepressant
treatment, this is considered ‘pseudo dementia’[11]
and, thus, is differentiated from AD. However, if
residual cognitive symptoms remain after effective
antidepressant treatment – as occurred in this case –
these cognitive symptoms can either be considered
residual symptoms of depression or, alternatively,
the behavioral and cognitive manifestations of a
separate disorder (dementia). In this patient, the
presence of transient auditory hallucinations and
other symptoms which were not present in her prior
episodes of depression suggest an underlying organic
brain disease; if this is the case, the current episode of
depression could be an indicator of prodromal AD.[12]
Two additional factors further complicated the
diagnosis. First, the patient had used reserpine
(which depletes monoamines and is often a cause
of depression) as a treatment for hypertension for
several years prior to admission. As is true for the
majority of such cases,[13] the depressive and anxiety
symptoms improved after stopping the reserpine (and
giving antidepressants). The failure to see substantial
concurrent cognitive improvement after stopping
reserpine could be to several reasons: there may be
an independent dementing process; the long-term
use of reserpine may have caused permanent brain
damage; or cognitive improvement may take much
longer than improvement in depression because it
takes a long time to fully restore depleted monoamine
reserves. Another complicating issue is the patient’s
lack of family support, something essential in the
management of geriatric depression.[14,15] Having no
children, her elderly husband was her sole care-giver;
her infrequent visits to the outpatient department
made it impossible to manage her complicated
condition appropriately and to observe the gradual
onset of cognitive impairment.
Currently popular treatments for patients suffering
from depression with cognitive dysfunction include
medication combined with psychological and physical
treatments such as repetitive transcranial magnetic
stimulation (rTMS) and modified electroconvulsive
therapy (MECT).[16] In this case, inpatient admission
was required to re-adjust her medications and
to provide psychotherapy and opportunities for
increasing her social activities. After 8 weeks of
treatment her depressive and anxiety symptoms had
improved significantly and her indifference and social
inactivity improved slightly, but there was relatively
little improvement in her cognitive impairments. We
conclude that long-term follow-up, probably for twoto-three years, will be needed to definitively determine
the primary cause (or causes) of her cognitive
symptoms.
Many elderly individuals who seek help from
health services suffer from both depression and
cognitive impairment. There are several important
lessons that this case highlights. (a) Elderly patients
with underlying cognitive impairment who experience
a depressive episode are more likely to develop
dementia[17] and have a poor response to medication.[18]
This complicates the diagnosis and management of
such patients and, thus, necessitates both a detailed
physical exam and a thorough neuropsychological
evaluation prior to the initiation of treatment. (b)
Longitudinal observation of the course of illness may
be needed to determine the correct diagnosis. A twoyear longitudinal study of 201 non-demented elderly
patients who experienced depression with cognitive
dysfunction found that 50 (25%) recovered, 30 (15%)
developed dementia, and 121 (60%) maintained some
cognitive dysfunction.[19] (c) Clinicians who treat elderly
patients must be aware of the cognitive effects of
commonly used psychoactive drugs, and they must
be up-to-date on the potential affective and cognitive
side-effects of the full range of medications used to
treat the physical illnesses experienced by elderly
individuals. A research study in the United States
estimated that there are 5.6 to 8.0 million people over
the age of 65 who misuse medications that often cause
cognitive impairment, including benzodiazepines,
anticholinergics, opioids, analgesics, hypnotics, and
antipsychotics. [20] (d) Family and social support is
important to the quality of life and social functioning
of all elderly individuals, but it is even more important
for elderly persons with depression or dementia.
These individuals need substantial help in managing
their condition beyond what can be provided by the
health care system. Part of the clinical assessment
of such individuals should include an assessment of
the types of social support available to the patient.
Subsequent treatment planning should include
provision of educational and psychological support to
• 113 •
primary care-givers, and, when necessary, involvement
of community-based social services to assist when the
available family resources are insufficient.
The authors declare that they have no conflict of
interest related to this manuscript.
Funding
Informed consent
The patient and her guardian signed an informed
consent form and agreed to the publication of this
case report.
Preparation of this report was supported by grants
from the Shanghai Clinical Center for Mental Disorders
(2014), by the National Key Clinical Disciplines program
at the Shanghai Mental Health Center (Office of
Medical Affairs, Ministry of Health, 2011-873; OMAMH, 2011-873), and by the Science and Technology
Commission of Shanghai’s Medical Guide Project
(No.15411961400).
Authors’ contribution
SZY participated in data collection and drafted the
manuscript. XSF and LX carried out the clinical
diagnosis and treatments. LX critically reviewed the
manuscript. All authors read and approved the final
manuscript.
难治性抑郁还是老年期痴呆 : 一例病例报告
石中永，肖世富，李霞
概述：抑郁伴痴呆在老年人中日益普遍。本报告描
述了一个 78 岁的女性患者，先前有过两次抑郁发作，
本次存在抑郁症状（动力缺乏和情感淡漠）和典型
的痴呆症状（记忆力和执行功能受损）。即使经过
详细的临床检查和神经心理测量，仍然难以明确诊
断是难治性抑郁症还是老年痴呆。经过 8 周的住院
治疗，更改了原先以利血平为主的降压药，调整抗
抑郁药并予心理治疗，患者的抑郁和焦虑症状改善，
但大多数认知症状仍然持续存在。在出院后 7 个月
的随访中，这些症状也没有变化。随后，她出现了
晚期乳腺癌并开始化疗，此时她的抑郁症状和认知
症状更加明显。我们认为，需要 2~3 年的随访才可
以确定认知症状是抑郁症的残留症状还是新出现的
痴呆表现（或两者皆是）。该病例表明对于同时有
抑郁症状和痴呆症状的老年患者，不仅需要详细的
临床检查和神经心理测试，而且要结合对治疗疗效
的长期评估才能明确诊断。
关键词：抑郁；痴呆；假性痴呆；病例报告；中国
References
1.
2.
3.
4.
5.
6.
7.
Hamilton M. A rating scale for depression. J Neurol
Neurosurg Psychiatry. 1960; 23(1): 56-62
Hamilton M. The assessment of anxiety states by rating. Br
J Med Psychol. 1959; 32(1): 50-55
Zhang MY, Katzman R, Salmon D, Jin H, Cai GJ, Wang ZY, et
al. The prevalence of dementia and Alzheimer’s disease
in Shanghai, China: impact of age, gender, and education.
Ann Neurol. 1990; 27(4): 428-437
Nasreddine ZS, Phillips NA, Bédirian V, Charbonneau
S, Whitehead V, Collin I, et al. The Montreal Cognitive
Assessment, MoCA: a brief screening tool for mild
cognitive impairment. J Am Geriatr Soc. 2005; 53(4):
695-699. doi: http://dx.doi.org /10.1111/j.15325415.2005.53221.x
National Cooperative Group for Wechsler Adult
Intelligence Scale-Revised. Wechsler Adult Intelligence
Scale-Revised. Acta Psychologica Sinica. 1983; 15(3): 362369.
Gong YX. [Wechsler Memory Scale – Revised in China].
Hunan: Hunan Medical University Publishing; 1989.
Chinese
Zhu ZQ, Ji JL, Xiao SF. [Key of Depression Treatment].
Jiangshu: Jiangsu Science and Technology Publishing
House; 2003. pp: 142-143. Chinese
8.
American Psychiatric Association. Diagnostic and
Statistical Manual of Mental Disorders, 5th ed. Arlington,
9.
Bhalla RK, Butters MA, Becker JT, Houck PR, Snitz BE,
Lopez OL, et al. Patterns of mild cognitive impairment
after treatment of depression in the elderly. Am J Geriatr
Psychiatry. 2009; 17(4): 308-316. doi: http://dx.doi.
org/10.1097/JGP.0b013e318190b8d8
10.
Morimoto SS, Alexopoulos GS. Cognitive deficits in
geriatric depression: clinical correlates and implications
for current and future treatment. Psychiatr Clin North
Am. 2013; 36(4): 517-531. doi: http://dx.doi.org/10.1016/
j.psc.2013.08.002
11.
Kang H, Zhao F, You L, Giorgetta CDV, Sarkhel S, Prakash
R. Pseudo-dementia: a neuropsychological review. Ann
Indian Acad Neurol. 2014; 17(2): 147-154. doi: http://
dx.doi.org/10.4103/0972-2327.132613
12.
Wang S, Blazer DG. Depression and cognition in the
elderly. Annu Rev Clin Psychol. 2015; 11: 331-360
13.
Baumeister AA, Hawkins MF, Uzelac SM. The myth of
reserpine-induced depression: role in the historical
development of the monoamine hypothesis. J Hist
Neurosci. 2003; 12(2): 207-220
• 114 •
14.
15.
16.
17.
Michael Ebert PT, Looen BN. [Diagnosis and Treatment
of Modern Mental Illness]. Sun XL, translator. Beijing:
People's Health Publishing House; 2002. p. 316. Chinese
Z h e n g Z , M a o J, S u n Y Y. [ I m p a c t o f t h e fa m i l y
structure on elderly depression]. Zhong Hua Xing
Wei Yi Xue Yu Nao Ke Xue Za Zhi. 2013; 22(11): 1016.
Chinese. doi: http://dx.chinadoi.cn/10.3760/cma.
j.issn.1674-6554.2013.11.018
Koenig AM, Butters MA. Cognition in late life depression:
treatment considerations. Curr Treat Options Psychiatry.
2014; 1: 1-14
Potter GG, Wagner HR, Burke JR, Plassman BL, WelshBohmer KA, Steffens DC. Neuropsychological predictors
of dementia in late-life major depressive disorder. Am
J Geriatr Psychiatry. 2013; 21(3): 297-306. doi: http://
dx.doi.org/10.1016/j.jagp.2012.12.009
18.
Sneed JR, Culang ME, Keilp JG, Rutherford BR, Devanand
DP, Roose SP. Antidepressant medication and executive
dysfunction: a deleterious interaction in late-life
depression. Am J Geriatr Psychiatry. 2010; 18(2): 128-135.
doi: http://dx.doi.org/10.1097/JGP.0b013e3181c796d2
19.
Steffens DC, McQuoid DR, Potter GG. Outcomes of
older cognitively impaired individuals with current
and past depression in the NCODE study. J Geriatr
Psychiatry Neurol. 2009; 22: 52-61. doi: http://dx.doi.
org/10.1177/0891988708328213
20.
Blank K. Older adults & substance use: new data highlight
concerns. Substance Abuse & Mental Health Service
Administration (SAMHSA); 2009
(received 2015-07-31; accepted 2015-12-15)
Zhongyong Shi graduated with a bachelor’s degree in clinical psychology from Henan University in
2013. She is currently a masters’ degree student in the Tenth People’s Hospital of Tongji University
on a clinical rotation at the Shanghai Mental Health Center. Her main research interest is biomarkers
of cognitive impairment.
• 115 •
•Biostatistics in psychiatry (32)•
Correlation and agreement: overview and clarification of
competing concepts and measures
Jinyuan LIU1, Wan TANG2, Guanqin CHEN1, Yin LU3,4, Changyong FENG1, Xin M. TU1,*
Summary: Agreement and correlation are widely-used concepts that assess the association between
variables. Although similar and related, they represent completely different notions of association.
Assessing agreement between variables assumes that the variables measure the same construct, while
correlation of variables can be assessed for variables that measure completely different constructs. This
conceptual difference requires the use of different statistical methods, and when assessing agreement
or correlation, the statistical method may vary depending on the distribution of the data and the interest
of the investigator. For example, the Pearson correlation, a popular measure of correlation between
continuous variables, is only informative when applied to variables that have linear relationships; it may be
non-informative or even misleading when applied to variables that are not linearly related. Likewise, the
intraclass correlation, a popular measure of agreement between continuous variables, may not provide
sufficient information for investigators if the nature of poor agreement is of interest. This report reviews the
concepts of agreement and correlation and discusses differences in the application of several commonly
used measures.
Keywords: concordance correlation; intraclass correlation; Kendall's tau; non-linear association;
Pearson's correlation; Spearman's rho
1. Introduction
Agreement and correlation are widely used concepts
in the medical literature. Both are used to indicate the
strength of association between variables of interest,
but they are conceptually distinct and, thus, require the
use of different statistics.
Correlation focuses on the association of changes
in two outcomes, outcomes that often measure quite
different constructs such as cancer and depression.
The Pearson correlation is the most popular measure
of the association between two continuous outcomes,
but it is only useful when measuring linear relationships
between variables. If the relationship is non-linear, the
Pearson correlation generally does not provide a good
indication of association between the variables. Another
problem is that using the standard interpretation
of Pearson correlation coefficients can, in some
circumstances, lead to incorrect conclusions.
Agreement, also known as reproducibility, is a
concept closely related to, but fundamentally different
1
Department of Biostatistics and Computational Biology, University of Rochester, Rochester, NY, USA
Department of Biostatistics & Bioinformatics, School of Public Health and Tropical Medicine, Tulane University, New Orleans, LA, USA
3
VA Cooperative Studies Program Palo Alto Coordinating Center, VA Palo Alto Health Care System, Palo Alto, CA, USA
4
Department of Biomedical Data Science, Stanford University, Stanford, CA, USA
2
*correspondence: Professor Xin M. Tu, Department of Biostatistics and Computational Biology, University of Rochester, 601 Elmwood Ave. Box 630,
CTSB 4.239, Rochester, NY 14642, USA.. E-mail: [email protected]
• 116 •
from, correlation. Like correlation, agreement also
assesses the relationships between outcomes of
interest, but, as the name indicates, the emphasis is
on the degree of concordance in the opinions between
two or more individuals or in the results between
two or more assessments of the variable of interest.
An example of agreement in mental health research
is the consensus between multiple clinicians about
the psychiatric diagnoses of a group of patients. In
biomedical sciences agreement can also include
measures of the reproducibility (i.e., reliability) of
a laboratory test result when repeated in the same
center or when conducted in multiple centers under
the same conditions. It is not sensible to speak of
agreement (reproducibility) between variables that
measure different constructs; so when measuring
the association between different variables – such as
weight and height – one can assess correlation but not
agreement. For continuous outcomes, the intraclass
correlation (ICC) is a popular measure of agreement.
Like the Pearson correlation, the ICC is an estimate of
the magnitude of the relationship between variables (in
this case, between multiple assessments of the same
variable). However, the ICC also takes into account rater
bias, the element that distinguishes agreement from
correlation; that is, good agreement (reproducibility)
not only requires good correlation, it also requires small
rater bias.
In this report, we provide an overview of popular
measures and statistical methods for assessing the two
different notations of association between variables. We
also clarify the key differences between the measures
and between the methods used to assess the measures.
We focus on continuous outcomes and assume all
variables are continuous unless stated otherwise.
2. Correlation measures
2.1 Pearson correlation
Consider a sample of n subjects and a bivariate
continuous outcome, (ui, vi), from each subject within
the sample (1≤i≤n). The Pearson correlation is the most
popular statistic for measuring the association between
the two variables ui and vi : [1]
!
p =
| ni = 1 ( u i - u-.) ( v i - v.)
|n
i=1
-) 2
( u i - u.
2
| n ( vi - v.)
i=1
n
n
1 |
- = 1 | ui , u.
v. =
vi ,
n i=1
n i=1
,
(1)
where u.(v.) denotes the sample mean of u i (v i ) The
Pearson correlation ⌒
p ranges between -1 and 1, with
1(-1) indicating perfect positive (negative) correlation
and 0 indicating no association between the variables.
As popular as it is, the Pearson correlation is only
appropriate for measuring correlation between ui and
vi when the two variables follow a linear relationship.
If the bivariate outcome (u i, v i) follows a non-linear
⌒
is not an informative measure and is
relationship, p
difficult to interpret.
To see this, let μ u (μ v ) and σ 2u (σ 2v ) denote the
(population) mean and (population) variance of
the variable u i ( v i ) . T h e Pe a rs o n co r re l at i o n i s
an estimate of the following product moment
correlation:
p = Corr (u i, v i) =
Cov (u i, v i)
Var (u i) Var (v i)
=
;
E ( u i - n u) ( v i - n v )
2
2
vu vv
E.
(2)
⌒
Unlike p
, which measures correlation between
u i and v i based on the sample, the product-moment
correlation p is the population-level correlation, which
cannot be calculated but is estimated by ⌒
p may
p . Thus, ⌒
also be referred to as the ‘sample product-moment
correlation’.
If u i and v i have a linear relationship, then
u i=avi + b + ε i, where a and b are some constants, and
ε i denotes random errors with mean 0 and variance
σ2ε . By centering u i (v i ) at its mean, we have: u i - μ u =
a(v-μ v )+ε i . It follows that σu2 =a 2σv2 +σε2. If ui and vi are
perfectly correlated, that is, σ 2ε =0, it follows from
Equation (2) that p=1 or (-1), depending on whether a is
positive or negative. Also, if ui and vi are uncorrelated,
or independent, that is, a=0, then p=0 and vice versa.
If u i and v i have a non-linear relationship, the
product moment correlation generally does not
provide an informative measure of correlation. The
example below shows that the Pearson correlation in
this case can be quite misleading.
Example 1. Suppose that u i and v i are perfectly
correlated and follow the non-linear relationship,
ui=vi9. Further, assume that vi follows a standard normal
distribution N(0,1) with mean 0 and variance 1. Then,
the product-moment correlation is:
10
p=
10
9
E (v i ) - E (v i ) E (v i )
9
Var (v i ) Var (v i)
=
10
E (v i )
18
2
9
E (v i ) - E (v i )
=
E (v i )
18
E (v i )
= 0.161 .
(3)
The poor association between u i and v i as indicated
by the product-moment correlation contradicts the
conceptual perfect correlation between the two
variables. Thus, the product-moment and its sample
counterpart, the Pearson correlation, generally do not
apply to non-linear relationships.
2.2 Spearman's Rho
Spearman's rho is also a popular measure of association.
Unlike the Pearson correlation, it also applies to
non-linear relationship, thereby addressing the
aforementioned limitation associated with the Pearson
correlation.
Let q i (r i ) denote the rankings of u i (v i ),(1 ≤ i ≤ n ).
Spearman's rho is defined as:
• 117 •
!
t =
.
| ni=1 (q i - q.) (r i - r .)
2
2
| ni=1 (q i - | in=1 (r i - q .)
r.)
n
1 |
-. =
q
n i = 1 qi,
Spearman rho is an estimate of the following population
Spearman rho:
,
(4)
ρ=12E[I(u j<u i)I(v k<v i)]-3, for all 1≤i <j<k ≤n .
n
1 |
=
r.
n i = 1 ri .
In Equation (7), E[I(u j <u i )I(v k <v i )] stands for the
mathematical expectation of I(u j <u i )I(v k <v i ) and
I(u j <u i ) (similarly I(v k <v i )) denotes an indicator with
I(u j <u i )=1(0) if u j <u i . It can be shown that ρ⌒ =1(-1) if
(u i,v i) are perfectly concordant (discordant) and vice
versa.
Note that the sample Spearman's rho in (4) is
referred to as Spearman's rho in the literature. Unlike
the Pearson correlation, there is no formal name for the
population Spearman's rho in (7). In general, the lack
of a formal name for the population version does not
cause confusion, since it is usually clear which one is
used within the context of a discussion. Like all statistics,
the population version of a statistic is called a parameter
in statistical lingo. The statistic and parameter serve
different purposes. For example, only the parameter can
be used in stating statistical hypotheses, such as the null
hypothesis, H:ρ=0, for testing whether the population
Spearman's rho is 0. Reported values of Spearman's rho
by studies are always the sample Spearman rho.
By comparing (1) and (4), it is clear that ρ⌒ is really the
Pearson correlation when applied to the rankings (q i , r i)
of the original variables (ui,vi). Since the rankings only
concern the ordering of the observations, relationships
among the rankings are always linear, regardless of
whether the original variables are linearly related. Thus,
Spearman's rho not only has the same interpretation as
the Pearson correlation, but also applies to non-linear
relationships.
The Spearman ρ⌒ ranges between -1 and 1, with 1
and -1 indicating perfect positive (negative) correlation;
when ⌒ρ=0 there is no association between the variables
ui and vi . If ⌒
ρ =1 then q i = r i, in which case,
(5)
u i <u j , v i <v j or u i >u j , v i >v j for all 1≤i <j ≤n .
If ⌒
ρ =-1, then q i=n-r i+1, in which case,
u i <u j , v i >v j or u i >u j , v i <v j for all 1≤i <j ≤n .
(7)
(6)
Any two pairs of bivariate outcomes (u i,vi) and (uj,vj)
that satisfy (5) or (6) are said to be concordant or
discordant; that is, u i and v i are either both larger
or both smaller than u j and v j . Thus, perfect positive
(negative) correlation by Spearman' rho corresponds to
perfect concordance (discordance); that is, concordant
(discordant) pairs (ui,vi) and (uj,vj) for all 1≤i <j ≤n .
2.3 Kendall's Tau
Another alternative for non-linear association is
Kendall's tau.[2] Like Spearman's rho, Kendall's tau also
exploits the concept of concordance and discordance
to derive a measure for bivariate outcomes. Unlike
Spearman's rho, it uses the notion of concordant
and discordant pairs directly in the definition of this
correlation measure.
Specifically, Kendall's τ (sample version) is defined as:
! nc - nd
x =
nt ,
Example 2. Table 1 shows 12 observations of the
bivariate outcome (u i,v i) as described in Example 1,
and the ranks associated with these observations. Note
that ui and vi are perfectly related, so their rankings are
identical; that is, q i = r i.
In this example the Pearson correlation p⌒ =0.531,
while Spearman’s ρ⌒ =1. Thus, only the Spearman rho
captures the perfect non-linear relationship between ui
and vi .
⌒
=0.531 has
Note that the Pearson correlation p
a higher upward bias than the product-moment
correlation p=0.161; this occurs due to the small sample
size, n=12. As sample size increases, ⌒
p becomes closer
to p, a property known as ‘consistency’ in statistics.
For example, we also simulated (ui,vi) with n=1000 and
obtained ⌒
p =0.173, much closer to p.
Like the Pearson correlation, the Spearman's rho
in (4) is a statistic based on a sample. This sample
nt =
1
n (n - 1) ,
2
(8)
n c = number of concordant pairs,
n d = number of discordant pairs.
1
In the above, n t = 2 n (n - 1) is the total number
of concordant and discordant pairs in the sample. If
n c =n t (n d =n t ), then ⌒
τ =1(-1) and vice versa. Also, if
there is no association between ui and vi , then nc
and nd should be close to each other and ⌒
τ should be
close to 0 (not exactly 0 due to sampling variability).
Table 1. A sample of 12 bivariate outcomes (ui,vi) simulated with u i = vi9 and vi from standard normal N (0,1).
ui
0.26
1.49
1.39
0.65
-0.49
-1.38
1.168
0.87
-0.96
2.15
-0.03
-1.08
vi
0
38.1
19.4
0.02
-0.002
-18.5
4.06
0.29
-0.68
971.6
0
-2.10
q i ( r i)
6
11
10
7
4
1
9
8
3
12
5
2
• 118 •
Thus, like Spearman's rho, ⌒
τ =1(-1) corresponds to
perfect concordance (discordance). A value of ⌒
τ close
to 0 indicates weaker or no association between the
variables ui and vi .
Like the Pearson and Spearman correlation, the
τ in (8) estimates the following
sample Kendall's ⌒
population parameter:
τ =2E[I(u i<u j)I(v i<v j)]-1, for all 1≤i <j ≤n .
Like its sample counterpart, τ also ranges between -1
and 1. If (5) holds true for all pairs (u i,v i) and (u j,v j),
then E[I(ui<uj)I(vi<vj)]=1 and τ=1. Likewise, if (6) holds
true for all pairs, then E[I(u i<u j)I(v i<v j)]=0 and ⌒
τ =-1.
Thus,τ = 1 (-1) corresponds to perfect concordance
(discordance). Finally, if ui and vi are independent, then
1
and τ = 0 . Thus,τ = 0 indicates
E 6I ^u i 1 u j h I ^ vi 1 vj h@ =
2
no association between ui and vi , and vice versa.
Example 3. Consider the data in Example 2. The sample
Kendall’s tau ⌒
τ =-1. Thus, like Spearman’s rho, Kendall’s
tau also provides a sensible measure of association for
non-linearly related variables.
3. Agreement and measures of agreement
Agreement, or reproducibility, is another widely used
concept for assessing the relationship among outcomes.
As indicated in the Introduction, unlike variables
considered in correlation analysis, variables considered
for agreement must measure the same construct.
Conversely, measures of correlation considered in
Section 2 generally do not apply to agreement.
Example 4. Consider two judges who rate each subject
from a study of 5 subjects sampled from a population of
interest using a scale from 1 to 10. Let ui and vi denote
the two judges' ratings on the ith subject (1<i<5).
Suppose that the judges' ratings from the subjects are
as follows:
(ui,vi) : (1,6), (2,7), (3,8), (4,9), (5,10).
Since u i and v i are linearly related, the Pearson
correlation can be applied, yielding ⌒
p =1, indicating
perfect correlation. However, the data clearly do not
indicate perfect agreement; in fact, the two judges
hardly agree with one another.
The poor agreement in this hypothetical example is
due to bias in judges' ratings. The mean ratings for the
two judges are 3 (for ui) and 8 (for vi ). Thus, despite the
perfect correlation between the ratings, the two judges
do not have good agreement because of bias in their
ratings of the subjects; either ui has downward or vi has
upward bias (or both).
The issue of bias does not apply to correlation
because the variables considered for correlation generally
measure different constructs and, thus, typically have
different means. For the Pearson correlation, the sample
means u. and v. are removed from the calculations of
the correlation in (1), thus, the Pearson correlation is
independent of differences between the (sample) means
of the variables being correlated.
3.1 Intraclass correlation
Intraclass correlation (ICC) is a popular measure of
agreement for continuous outcomes. Like the Pearson
correlation, the ICC requires a linear relationship
between the variables. However, it differs from the
Pearson correlation in one key respect; the ICC also takes
into account differences in the means of the measures
being considered. In addition, the ICC can be applied to
situations where there are three or more separate raters.
Consider a study with n subjects and assume each
subject is rated by a different group of K judges. Let
yik denote the rating of the ith subject by the kth judge
(1 ≤ i ≤ n , 1 ≤ k ≤ K ). The ICC is defined based on the
following linear mixed-effects model:[3-5]
yik=μ+β i+ ε ik , 1≤k≤K, 1≤i≤n,
(9)
βi ~ N (0, σβ2 ), εik ~ N (0, σ2 ).
In the above model, the fixed effect μ is the (population)
mean rating of the study population over all possible
K judges from the population of judges; that is, the
random effect or latent variable. β i represents the
difference between the mean rating of the ith subject
and the mean rating of the study population μ. Thus,
the sum u+β i represents the mean rating of the i th
subject. The intraclass correlation (ICC) is defined as
v
2
the variance ratio, pICC = 2 b 2 ,, of the variance σβ2 of the
vb + v
mean rating of the subjects (u+βi) to the total variance
consisting of σβ2 plus the variance σ2 of the judges.
If there are only two judges (K=2), then under
the linear mixed-effects model in (9) the productmoment correlation between yi1 and yi2 is the same as
the ICC; that is, Corr (y i1, y i2) = v
2
vb
2
b
+v
2
. Moreover, yi1 and
yi2 have the same mean (μ) and variance (σ2 ). Thus, in
this special case, the ICC is the same as the productmoment correlation (pICC= p). Note that this result is
not a contradiction to the data in Example 4, since ui
and vi do not have the same mean and thus the linear
mixed-effects model in (9) does not apply to the data
and the ICC no longer serves its intended purpose in
this case. However, since differences in means between
judges’ ratings decrease the ICC, this agreement index
may still be applied in this situation to indicate poorer
agreement. Follow-up analyses are necessary to
determine whether poor agreement is due to bias or
large variability or both between the judges.
Example 5. Consider again Example 4 and let yi1=u i
and y i1=v i. By fitting the model in (9) to the data,
we obtain estimates σ⌒ β2 = 0 and σ⌒ 2 =9.167. Thus, the
(sample) ICC based on the data is ⌒
pICC =0, which is quite
different from the Pearson correlation. Although the
judges' ratings are perfectly correlated, agreement
between the judges is extremely poor.
• 119 •
Note that p⌒ICC is not a valid measure of agreement
between yi1 and yi2 for the data in Example 5, since
the assumption of a common mean between yi1 and
yi2 is not met by the data. However, it is precisely this
assumption that makes p⌒ICC totally different from the
Pearson correlation ⌒
p =(1). We may revise the model in (9)
to account for the bias in the judges' ratings to consider:
yik=μk+βi+εik , 1≤k≤K, 1≤i≤n,
(10)
βi ~ N (0, σβ2 ), εik ~ N (0, σ2 ),
where the added fixed-effect μ k accounts for the
difference between the two judges. By fitting the above
⌒
σ β2 =1.256, ⌒
σ 2 =0, μ
model, we obtain estimates ⌒
1=3 and
⌒
μ 2 =5. Once accounting for bias, the two judges have
perfect agreement. The model in (10) also provides
⌒
⌒2
mean ratings μ
K for the judges. The positive estimate σ β
describes the variability among the subjects. Although
the correct model for the data, the ICC calculated from
the model in (10) no longer has the interpretation as a
!2
measure of agreement. In fact, ! 2 ! 2 = 1, the same
vb
vb + v
as the Pearson correlation ⌒
p =1 as we have calculated in
Example 4.
Note since pICC≥0 we can either reverse code some
of the judges' ratings or use a different index, such as
the concordance correlation, discussed below.
3.2 Concordance correlation
The concordance correlation (CCC) is another measure
of agreement which, unlike the ICC, does not assume a
common mean for judges' ratings at the outset, so it can
be used to assess both the level of agreement and the
level of disagreement. However, a major limitation of
the CCC is that it only applies to two judges at a time.
Consider a study with n subjects and assume
each subject is rated by a different group of two
judges. Let y ik again denote the rating of the i th
subject by the kth judge (1≤i≤n, 1≤k≤2). Let μ k= E(yik)
and σk2 =Var(yik), denoting the mean and variance of yik,
and σ12=Cov(yi1, yi2), denoting the covariance between
yi1 and yi2. The CCC is defined as:[6]
2v 12
Pccc = 2
(11)
2
2 .
v 1 + v 2 + (n 1 - n 2)
Unlike the ICC, no statistical model is assumed in the
definition of pCCC. Further, the two judges can come
from two different populations of judges with different
means and variances.
The CCC pCCC has a nice decomposition, pCCC=pCb,
where p is the product-moment correlation in (2) and Cb
is called the bias correction factor given by:
2
Cb = v
(12)
v 2 (n 1 - n 2) 2 .
1
v 2 + v 1 + v 1v2
It can be shown that pCCC=1(-1) if and only if p=1(-1),
μ1=μ2 and σ12 = σ 22 .[6] Thus, pCCC=1(-1) if and only if yi1 =
yi2(yi1=-yi2), that is, when there is perfect agreement
(disagreement). The bias correction factor Cb(0≤Cb≤ 1 ) in
(12) assesses the level of bias, with smaller Cb indicating
larger bias. Thus, unlike the ICC, poor agreement can
result from low correlation (small p) or large bias
(small Cb).
Example 6. Consider again Example 5. The (sample)
mean and variance of yi1, and the (sample) correlation
between yi1 and yi2 are given by: μ⌒1=3, ⌒
σ 12 =2.5,
μ 2 =8, ⌒
⌒2
⌒
σ 2 =2.5 and σ 12 =1. Thus, it follows from (11) that
!
2 v 12
!
p ccc = !2 !2 ! !
= 0.0533. .We can also
v 1 + v 2 + ( n 1 - n 2) 2
⌒
obtain p
CCC by using the decomposition result, which in
our case yields ⌒
p CCC = ⌒
p⌒
p =1, ⌒
Cb=0.0533.
C b= 0.0533 and ⌒
Note that unlike correlation the issue of linear
versus non-linear association does not arise when
assessing agreement. This is because good agreement
requires an approximate linear relationship between the
outcomes. For example, in the case of two raters, good
agreement requires that yi1 and yi2 are close to each
other, such as yi1 = yi2 in the case of perfect agreement.
4. Discussion
We discussed the concepts of agreement and correlation
and described various measures that can be used to
assess the relationships among variables of interest. We
focused on the measures and methods for continuous
outcomes. For non-continuous outcomes, different
methods must be applied. For example, for categorical
outcomes a different version of Kendall's tau, known as
Kendall's tau b can be used for assessing correlation and
Kappa can be used for assessing agreement.[7]
Funding
The work was supported in part by a grant (GM108337)
from the National Institutes of Health and the National
Science Foundation (Tang and Tu) and a pilot grant
(UR-CTSI GR500208) from the Clinical and Translational
Sciences Institute at the University of Rochester
Medical Center (Feng and Tu).
The authors report no conflict of interest.
All authors worked together on this manuscript. In
particular, JYL, WT and XMT made major contributions
to the section on correlation, GQC, YL and CYF made
major contributions to the section on agreement, and
JYL and XMT drafted and finalized the manuscript. All
authors read and approved the final manuscript.
• 120 •
相关性和一致性：这对相仿概念和测量方法的回顾与阐明
Liu JY, Tang W, Chen GQ, Lu Y, Feng CY, Tu XM
概述：一致性 (agreement) 和相关性 (correlation) 是两
个广泛使用的概念，用来评估变量之间的关联。虽然
二者相似且相关，但是它们代表关联完全不同的概念。
评估变量之间的一致性假设变量测量的是相同的结构，
而在变量测量完全不同的结构时也可以评估它们之间
的相关性。这种概念上的差异就要求使用不同的统计
方法，并且当评估一致性或相关性时，统计方法根据
数据的分布和研究者的兴趣可能会有所不同。例如，
Pearson 相关性，作为评估连续变量之间相关性的一种
普遍测量方法，只有用于符合线性关系的变量时才能
提供有用的信息；当用于不符合线性关系的变量时就
无法提供准确信息甚至会产生误导。同样地，内部相
关性，作为一种评估连续变量之间一致性的常用方法，
如果一致性不好的实质正好是研究兴趣所在，那么该
测量就不能为研究者提供充分的信息。本报告回顾了
一致性和相关性的概念，并讨论了几种常用方法在应
用中的差异。
关键词：积差相关性，内部一致性，Kendall's tau，非
线性相关，Pearson's 相关性，Spearman's rho
References
1. Stigler SM. Francis Galton's Account of the Invention of
Correlation. Statist Sci. 1989; 4(2): 73-79. doi: http://dx.doi.
org/10.1214/ss/1177012580
4. McGraw KO, Wong SP. Forming inferences about some
intraclass correlation coefficients. Psychol Methods. 1996; 1:
30-46. doi: http://dx.doi.org/10.1037/1082-989X.1.4.390
2. Kowalski J, Tu XM. Modern Applied U Statistics. New York:
Wiley; 2007
5. Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing
rater reliability. Psychol Bull. 1979; 86(2): 420-428
3. Lu N, Chen T, Wu P, Gunzler D, Zhang H, He H, et al. Functional
response models for intraclass correlation coefficients.
Applied Statistics. 2014; 41: 2539-2556. doi: http://dx.doi.org
/10.1080/02664763.2014.920780
6. Lin LI. A concordance correlation coefficient to evaluate
reproducibility. Biometrics. 1989; 45(1): 255-268
7. Tang W, He H, Tu XM. Applied Categorical and Count Data
Analysis. Boca Raton, FL: Chapman & Hall/CRC; 2012
Ms. Jinyuan Liu obtained her bachelor’s of science degree in statistics from Nanjing University
of Posts and Telecommunications in 2015. She is currently a master's student in the Department
of Biostatistics and Computational Biology at the University of Rochester in New York, USA. Her
research interests include categorical data analysis, machine learning, and social networks.
• A1 •
THE AMERICAN JOURNAL OF PSYCHIATRY
Volume 173 • Number 2 • February 2016
EDITORIALS
99
Maternal Defense Mechanisms Influence
Infant Development
J. Christopher Perry
101 Mothers, Babies, Depression, and Medications:
Understanding the Complex Interplay of Illness
and Treatment on Neonatal Symptoms
Linda H. Chaudron
103 Community Evidence of Clozapine’s
Effectiveness
Stephen R. Marder
105 Dose Response for SSRIs
Madhukar H. Trivedi
TREATMENT IN PSYCHIATRY
107 Clinical Experience With High-Dosage
Pramipexole in Patients With TreatmentResistant Depressive Episodes in Unipolar and
Bipolar Depression
Jan Fawcett,
PERSPECTIVES IN GLOBAL MENTAL HEALTH
112 A 39-Year-Old “Adultolescent”: Understanding
Social Withdrawal in Japan
Takahiro A. Kato
REVIEWS AND OVERVIEWS
117 Risk of Postpartum Relapse in Bipolar Disorder
and Postpartum Psychosis: A Systematic
Review and Meta-Analysis
Richard Wesseloo
128 The Sequential Integration of Pharmacotherapy
and Psychotherapy in the Treatment of Major
Depressive Disorder: A Meta-Analysis of the
Sequential Model and a Critical Review of the
Literature
Jenny Guidi
ARTICLES
138 Defense Mechanisms of Pregnant Mothers
Predict Attachment Security, Social-Emotional
Competence, and Behavior Problems in Their
Toddlers
John H. Porcerelli
147 The Roles of Maternal Depression, Serotonin
Reuptake Inhibitor Treatment, and
Concomitant Benzodiazepine Use on Infant
Neurobehavioral Functioning Over the First
Postnatal Month
Amy L. Salisbury
158 Heritability of Perinatal Depression and
Genetic Overlap With Nonperinatal Depression
Alexander Viktorin
166 Comparative Effectiveness of Clozapine and
Standard Antipsychotic Treatment in Adults
With Schizophrenia
T. Scott Stroup
174 Effect of Attention Training on Attention Bias
Variability and PTSD Symptoms: Randomized
Controlled Trials in Israeli and U.S. Combat
Veterans
Ewgeni Jakubovski
184 Longitudinal Psychiatric Symptoms in
Prodromal Huntington’s Disease: A Decade of
Data
Eric A. Epping
LETTERS TO THE EDITOR
193 Outcome Variation in the Randomized Trial
of Cognitive-Behavioral Therapy Versus Light
Therapy for Seasonal Affective Disorder
Arthur Rifkin
193 Response to Rifkin
Kelly J. Rohan
193 Reflections on “Addressing Patients’ Psychic
Pain”
Jon E. Gudeman
194 Twitter Article Mentions and Citations: An
Exploratory Analysis of Publications in the
American Journal of Psychiatry
Daniel S. Quintana, Nhat Trung Doan
BOOK FORUM
195 Global Mental Health: Anthropological
Perspectives
George S. Alexopoulos
196 DSM-5® Handbook on the Cultural Formulation
Interview
Rajiv Radhakrishnan
197 Oxford Textbook of Correctional Psychiatry
Peter Ash
198 Books Received
• A2 •
THE AMERICAN JOURNAL OF PSYCHIATRY
Volume 173 • Number 3 • March 2016
EDITORIALS
205 Isn’t Your Staff Trained To Manage My Mother?
Martin Steinberg
208 Evidence-Based Pregnancy Registries: Good for
Babies and Their Mothers
Vivien K. Burt
211 A New Option for Treating Bipolar I Depression
Holly A. Swartz, Joseph T. Tasosa
213 Dissecting the Brain Mechanisms of Violence
Robert Freedman, Robert Michels
CLINICAL CASE CONFERENCE
215 “Jinn Possession” and Delirious Mania in a
Pakistani Woman
Qurat ul ain Khan, Aisha Sanober
PERSPECTIVES IN GLOBAL MENTAL HEALTH
219 Displaced Iraqi Families in Kurdistan: Strangers
in a Strange Land
Rami Bou Khalil
REVIEWS AND OVERVIEWS
211 Post-Stroke Depression: A Review
Robert G. Robinson, Ricardo E. Jorge
232 A Selective Review of Cerebral Abnormalities
in Patients With First-Episode Schizophrenia
Before and After Treatment
Qiyong Gong
ARTICLES
244 Outcomes One and Two Winters Following
Cognitive-Behavioral Therapy or Light Therapy
for Seasonal Affective Disorder
Kelly J. Rohan
252 Impact of Antipsychotic Review and
Nonpharmacological Intervention on
Antipsychotic Use, Neuropsychiatric
Symptoms, and Mortality in People With
Dementia Living in Nursing Homes: A Factorial
Cluster-Randomized Controlled Trial by
the Well-Being and Health for People With
Dementia (WHELD) Program
Clive Ballard
263 Reproductive Safety of Second-Generation
Antipsychotics: Current Data From the
Massachusetts General Hospital National
Pregnancy Registry for Atypical Antipsychotics
Lee S. Cohen
271 An 8-Week Randomized, Double-Blind,
Placebo-Controlled Evaluation of the Safety
and Efficacy of Cariprazine in Patients With
Bipolar I Depression
Suresh Durgam
282 Neural Correlates of the Propensity for
Retaliatory Behavior in Youths With Disruptive
Behavior Disorders
Stuart F. White
291 Medial Prefrontal Aberrations in Major
Depressive Disorder Revealed by
Cytoarchitectonically Informed Voxel-Based
Morphometry
Sebastian Bludau
LETTERS TO THE EDITOR
299 Gene-Environment Interaction in Youth
Depression: Differential Susceptibility?
Eric M. Plakun
299 Response to Plakun: Addressing Differential
Susceptibility With Regard to GeneEnvironment Interaction in Youth Depression
Thiago Botter-Maio Rocha
300 Reflections on “Emil Kraepelin: Icon and
Reality”
Rael D. Strous
301 Response to Strous et al.: A Focus on
Kraepelin’s Clinical Research Methodology
Kenneth S. Kendler, Eric J. Engstrom
302 Going Beyond Finding the “Lesion”: A
Path for Maturation of Neuroimaging
Simon B. Eickhoff, Amit Etkin
BOOK FORUM
304 Psychiatric Polarities: Methodology and
Practice
Constantine G. Lyketsos, M.D., M.H.S

shanghai archives of psychiatry

Transcription

Similar documents

China 2006 July 7 - 18 Beijing, Xi`an, Shanghai and Hong Kong

PDF (Author`s version) - OATAO (Open Archive Toulouse Archive

Depression 2 - Tricia Worthington

Fact behind Building Collapse in Shanghai

10 Days only $2195 Departs: April 15,2000 B eijing • Xi`an • shanghai

Emotional Maltreatment

Alison Brettle - CILIP Conference

Sugar cane juice machine

Shanghai - Princess Cruises

Business, Capital, CEO, Civil - stylus | a project by ann hamilton