Full Terms & Conditions of access and use can be found athttp://www.tandfonline.com/action/journalInformation?journalCode=hbsm20

Download by: [George Washington University], [Alicia Yeung] Date: 28 September 2016, At: 07:06

Behavioral Sleep Medicine

ISSN: 1540-2002 (Print) 1540-2010 (Online) Journal homepage: http://www.tandfonline.com/loi/hbsm20

The Effects of Tai Chi on Sleep Quality in ChineseAmerican Patients With Major DepressiveDisorder: A Pilot Study

Yan Ma, Alicia Yeung, Albert C. Yang, Chung-Kang Peng, Alisabet Clain,Jonathan Alpert, Maurizio Fava & Albert S. Yeung

To cite this article: Yan Ma, Alicia Yeung, Albert C. Yang, Chung-Kang Peng, Alisabet Clain,Jonathan Alpert, Maurizio Fava & Albert S. Yeung (2016): The Effects of Tai Chi on Sleep Qualityin Chinese American Patients With Major Depressive Disorder: A Pilot Study, Behavioral SleepMedicine, DOI: 10.1080/15402002.2016.1228643

To link to this article: http://dx.doi.org/10.1080/15402002.2016.1228643

Published online: 27 Sep 2016.

Submit your article to this journal

View related articles

View Crossmark data

The Effects of Tai Chi on Sleep Quality in ChineseAmerican Patients With Major Depressive Disorder:

A Pilot Study

Yan MaDivision of Interdisciplinary Medicine and Biotechnology, Beth Israel Deaconess Medical

Center, Harvard Medical School, Boston, MassachusettsSleep Center Eye Hospital, China Academy of Chinese Medical Sciences, Beijing, China

Alicia YeungGeorge Washington University, Washington, DC

Albert C. YangDivision of Interdisciplinary Medicine and Biotechnology, Beth Israel Deaconess Medical

Center, Harvard Medical School, Boston, MassachusettsDepartment of Psychiatry, Taipei Veterans General Hospital, Taipei City, Taiwan

School of Medicine, National Yang-Ming University, Taipei City, Taiwan

Chung-Kang PengDivision of Interdisciplinary Medicine and Biotechnology, Beth Israel Deaconess Medical

Center, Harvard Medical School, Boston, Massachusetts

Alisabet Clain, Jonathan Alpert, and Maurizio FavaDepression Clinical and Research Program, Massachusetts General Hospital, Harvard Medical

School, Boston, Massachusetts

Chung-Kang Peng, PhD, is a co-inventor of the sleep spectrogram technique, which is patented and licensed toEmbla, Inc. by the Beth Israel Deaconess Medical Center.

Correspondence should be addressed to Yan Ma, Division of Interdisciplinary Medicine and Biotechnology, Beth IsraelDeaconess Medical Center, Harvard Medical School, 330 Brookline Ave, Boston, MA 02215. Email: dr.yan.ma@gmail.comand Albert S. Yeung, Massachusetts General Hospital, Depression Clinical Research Program, 1 Bowdoin Square, 6th Floor,Boston, MA 02114. Email: ayeung@partners.org

Behavioral Sleep Medicine, 00:1–17, 2016Copyright © Taylor & Francis Group, LLCISSN: 1540-2002 print/1540-2010 onlineDOI: 10.1080/15402002.2016.1228643

1

Albert S. YeungDepression Clinical and Research Program, Massachusetts General Hospital, Harvard Medical

School, Boston, MassachusettsDepartment of Behavioral Health, South Cove Community Health Center, Boston,

Massachusetts

Objective: This pilot study evaluated the effects of Tai Chi training on sleep quality (primaryoutcomes), and depression and social functioning levels (secondary outcomes) among patientswith depression. Participants: Sixteen depressed Chinese patients. Methods: Participants received1-hr Tai Chi training sessions 2 times per week for 10 weeks. Patients’ subjective sleep qualityratings, objective sleep quality measurements, and depression and social functioning levels weremeasured before, during, and after the intervention. Results: Sleep quality and depression outcomesimproved significantly. Patients reported improved Pittsburgh Sleep Quality Index (PSQI) scores (9.6± 3.3 to 6.6 ± 5.2, p = 0.016), and cardiopulmonary coupling (CPC) analysis of electrocardiogram(ECG) showed decreased stable sleep onset latency (75.7 ± 100.6 to 20.9 ± 18.0, p = 0.014),increased stable sleep percentages (31.5 ± 18.7 to 46.3 ± 16.9, p = 0.016), and decreased unstablesleep percentages (45.3 ± 20.1 to 30.6 ± 16.5, p = 0.003). Patients also reported decreased HamiltonRating Scale for Depression (HAM-D-17; 20.1 ± 3.7 to 7.8 ± 5.9, p < 0.001) and Beck DepressionInventory (BDI) scores (22.3 ± 9.1 to 11.1 ± 10.6, p = 0.006). Significant correlations were foundbetween the changes in subjective sleep assessments ΔPSQI and ΔHAM-D-17 (r = 0.6, p = 0.014),and ΔPSQI and ΔBDI (r = 0.62, p = 0.010). Correlations between changes in objective sleepmeasurements and changes in depression symptoms were low and not significant. Conclusions:Tai Chi training improved sleep quality and mood symptoms among depressed patients.

Insomnia, a sleep disorder that includes trouble falling asleep, staying asleep, or waking up tooearly, resulting in daytime impairment, is a common complaint among patients in primary careand behavioral health clinics (Punnoose, Golub, & Burke, 2012). Insomnia is also one of themost common symptoms of major depressive disorder (MDD; (American PsychiatricAssociation, 2013; Lai et al., 2014; Soehner, Kaplan, & Harvey, 2014). Pretreatment insomniaamong patients with depression predicts poorer outcomes of antidepressant treatment (Sung,Wisniewski, Luther, Trivedi, & Rush, 2015). Full recovery or remission from depression shouldbe the goal of depression treatment (Bakish, 2001; Kennedy, 2002) because the presence ofresidual depressive symptoms is associated with poorer depression outcomes (O’Brien et al.,2011), higher risk of suicide (Agargun et al., 1997; Don Richardson, Cyr, Nelson, Elhai, &Sareen, 2014; Li, Lam, Chan, Yu, & Wing, 2012), and greater risk of recurrent depression (Cole& Dendukuri, 2003; Perlis, Giles, Buysse, Tu, & Kupfer, 1997). As one of the most commonresidual symptoms in patients treated for depression (Carney, Segal, Edinger, & Krystal, 2007;Kanai et al., 2003; Pigeon et al., 2009), insomnia makes achieving a full recovery difficult(Troxel et al., 2012).

Electroencephalogram (EEG) based polysomnography (PSG) has been the established tech-nology for studying sleep quality. Sleep impairment in patients with MDD is characterized bysleep fragmentation, diminished slow-wave sleep, and altered rapid eye movement (REM) sleep(Cheng et al, 2015; Emslie, Rush, Weinberg, Rintelmann, & Roffwarg, 1990; Kudlow, Cha,Lam, & McIntyre, 2013). However, the use of EEG for sleep monitoring has many limitations.

2 MA ET AL.

One issue is that EEG only focuses on brain waves while sleep is a complex physiologicalprocess that involves all of the body’s organs and systems (Bianchi & Thomas, 2013). Recenttechnological advances allow the use of mobile devices to monitor physiological parameters,including electrocardiogram (ECG), in an inexpensive and continuous manner. Cardiopulmonarycoupling (CPC) analysis of continuous ECG data has been developed as a convenient alternativeto PSG in measuring sleep quality (Thomas et al., 2007; Thomas, Mietus, Peng, & Goldberger,2005). Yang and colleagues (Yang et al., 2011) compared sleep quality in depressed patients andhealthy controls using CPC analysis of sleep measurements and concluded that CPC analysis canbe used to quantify sleep quality or stability and to objectively evaluate the severity of insomniain patients with MDD. These studies have provided the basis for comprehensive assessment ofsleep using inexpensive mobile devices in the management of depression.

In the treatment of depression, pharmacological therapies using second-generation antidepres-sant treatments are the mainstay. However, there are significant side effects associated withantidepressants (Alberti, Chiesa, Andrisano, & Serretti, 2015; Hasnain, Vieweg, & Hollett,2012; Kikuchi, Suzuki, Uchida, Watanabe, & Mimura, 2013). Cognitive behavioral therapy andmind-body approaches are playing an increasing role in the treatment of depression and insomniaassociated with depression (Frame &Alexander, 2013; Geiger-Brown et al., 2014;Watanabe et al.,2015). Tai Chi is a comprehensive mind-body exercise originating from China, which is enjoyinggrowing popularity in theWest. It involves mild aerobic exercise, breathing techniques, attentionalfocus, mindfulness training, relaxation, postural control, along with choreographed movementpatterns. According to a 2007 National Health Interview Survey, 2.3 million Americans havepracticed Tai Chi in the past 12 months (Clarke, Black, Stussman, Barnes, & Nahin, 2015;Komelski, Miyazaki, & Blieszner, 2012). Preliminary research has shown beneficial effects ofTai Chi on a range of psychological well-being measures including mood, anxiety, general stressmanagement, self-esteem, and quality of life in varied populations (Ma, Sun, & Peng, 2014; Ma,Zhou, Fan, & Sun, 2016). Recent studies have provided growing evidence to suggest the efficacyof Tai Chi for patients with MDD (Cho, 2008; Chou et al., 2004; Lavretsky et al., 2011; Yeunget al., 2012) and those who suffer from sleep disturbances (Field, Diego, Delgado, & Medina,2013; Irwin et al., 2014; Larkey et al., 2015). To date, there are no studies that specifically evaluatethe effects of Tai Chi on quality of sleep among patients with MDD. In this study, we hypothesizethat patients’ subjective rating and objective sleep quality from CPC analysis of continuous ECGdata will show that Tai Chi improves sleep quality (primary hypothesis) and depression and socialfunctioning (secondary hypothesis) in patients with MDD.

Interested Chinese Americans from a community health center were prescreened for MDDusing the Patient Health Questionnaire (PHQ-9). Those with higher scores of PHQ-9 (≥ 10) wereinterviewed to confirm the diagnosis of MDD. Subjects with Hamilton Rating Scale forDepression (HAM-D-17) (Laenen, Alonso, Molenberghs, Vangeneugden, & Mallinckrodt,2009; Montgomery & Asberg, 1979) scores between 14 and 24 were selected. These patientswere considered to have moderate levels of depression that required intervention. Patients withsevere depression (HAM-D-17>24) were not included in this study due to ethical concerns as TaiChi is not considered an established treatment for MDD. The primary outcomes were patients’subjective sleep ratings and objective sleep quality measurements assessed by cardiopulmonarycoupling (CPC) (Thomas et al., 2005) analysis of ECG data collected using an ambulatorymonitoring device. Secondary outcomes were patients’ depression (measured by HAM-D-17,Clinical Global Impression [CGI] (Guy, 1976), and social functioning (measured by the Short

TAI CHI FOR SLEEP IN MDD PATIENTS 3

Form Health Survey (SF-36) (McHorney, Ware, Raczek, & The, 1993; Ware, Sherbourne, &The, 1992) levels.

METHODS

Materials and Study Design

This is a single-group, pre–post assessment of the effects of Tai Chi on sleep quality amongpatients with depression. This pilot study targeted Chinese patients with MDD. The study wasconducted between January 1 and November 30, 2014, at South Cove Community HealthCenter, a federally funded community health center in Boston, which predominantly servesChinese Americans.

Inclusion and Exclusion Criteria

Inclusion criteria included (a) self-identification as being of Chinese ethnicity and fluent inMandarin or Cantonese, (b) 18–65 years of age, (c) satisfy DSM-IV-R criteria for MDD, (d)have a baseline score of 14–24 on the HAM-D-17, and (e) have had no regular (definedas � 3times per week for � 2 months) Tai Chi training and practice or other forms of mind-bodyintervention in the past 6 months.

Exclusion criteria included (a) primary psychiatric diagnosis other than MDD, (b) history ofpsychosis, mania, or severe cluster B personality disorders, active alcohol or substance abuse ordependency disorders in the past 6 months, (c) unstable medical conditions as judged byinvestigators, (d) usage of or plans to use confounding treatments, including antidepressantsand CAM treatments thought to have beneficial effects on mood, such as St. John’s Wort,S-adenosyl methionine (SAMe), omega-3 fatty acids, light therapy, conventional psychotherapy,mind-body interventions (e.g., Qigong, mindfulness training, muscle relaxation training, etc.),(e) current active suicidal or self-injurious potential necessitating immediate treatment, (f)current pregnancy, (g) metallic implants, (h) claustrophobia, and (i) patients who have atrialfibrillation or an implanted pacemaker.

Subject Recruitment

Participants in this study were recruited through advertisement, referral by South Cove’s primarycare and mental health clinicians, and routine depression screening at South Cove’s primary careclinics. Since Tai Chi was taught in Chinese languages in this study, only participants who spokefluent Chinese were enrolled to ensure understanding, and to encourage social interaction andmutual support in class. The study was approved by the Institutional Review Board (IRB) of theMassachusetts General Hospital.

Potential subjects received a phone triage by our bilingual research staff using an IRB-approved protocol that includes both a study-specific phone screen questionnaire and thePHQ-9. If a patient was fluent in Mandarin or Cantonese Chinese, aged 18–65, had not beenpracticing Tai Chi regularly, and had a PHQ-9 score of 10 or above, the potential subject wasscheduled for a screening visit. During the screening visit, a bilingual investigator obtainedwritten consent from the subject, and conducted interviews. Screening visits occurred within

4 MA ET AL.

four weeks prior to the start date of the intervention. A psychiatrist interviewed the subject usinginstruments including the Chinese Bilingual version of the semistructured psychiatric interview(CB-SCID-I/P) to assess the presence of major DSM-IV-R psychiatric disorders (Nietzel &Wakefield, 1996), and the 17-item Hamilton Rating Scale for Depression (HAM-D-17) todetermine the eligibility of the patient. A urine pregnancy test was performed on female subjectswith child-bearing potential. Female participants who were over 60 years of age and self-reported as being postmenopausal were not required to complete the pregnancy test.

Tai Chi Intervention

Classes were conducted by a Tai Chi master who had more than 25 years of training experience. TaiChi participants received 1-hr Tai Chi training sessions 2 times per week for 10weeks. The instructorfollowed a standardized protocol, which included the first 24 of the traditional 108 movements ofYang-style Tai Chi. Participants were asked to practice Tai Chi at home with an instructional DVD atleast 3 times per week and 30 min each time. At the end of 10 weeks, the participants were expectedto be able to practice the 24 basic movements on their own. Tai Chi exercise logs were given toparticipants every week to record practice compliance and adverse events.

Scales and Questionnaires

Outcome measures were assessed at baseline, week 5, and week 10. At each assessment,participants were administered the Hamilton Rating Scale for Depression (HAM-D-17), BeckDepression Inventory (BDI), the Clinical Global Impression–Severity (CGI-S) and Improvement(CGI-I), the Short Form Health Survey (SF-36), the Multidimensional Scale of Perceived SocialSupport (MSPSS), Mindful Attention Awareness Scale (MAAS), and Pittsburgh Sleep QualityIndex (PSQI).

Hamilton Rating Scale for Depression (HAM-D-17)

HAM-D-17 is a clinician-administered interview scale that provides an indication of depres-sion, and acts as a guide to evaluate recovery. It has exhibited well-documented reliability andvalidity (Laenen et al., 2009; Montgomery & Asberg, 1979). HAM-D-17 assesses the presenceand severity of 17 symptoms of depression.

Beck Depression Inventory (BDI)

As a widely used psychometric test for measuring depression severity, BDI (Beck, Steer, Ball,& Ranieri, 1996; Beck, Ward, Mendelson, Mock, & Erbaugh, 1961) contains 21 questions aboutthe subject’s feeling in the previous week. Each question has a set of at least four possibleresponses, ranging in intensity. A value of 0 to 3 is assigned for each answer, and higher totalscores indicate more severe depressive symptoms.

Clinical Global Impression–Severity (CGI-S) and Improvement (CGI-I)

The CGI rating scales are commonly used for symptom severity and treatment response intreatment studies of patients with mental disorders (Guy, 1976). The CGI-S measures the current

TAI CHI FOR SLEEP IN MDD PATIENTS 5

condition of the patient, as judged by the clinician, on a scale of 1–7 (higher scores indicatemore severe depressive symptoms); and the CGI-I measures the degree of improvement, asjudged by the clinician, since the start of treatment on a scale of 1–7 (higher scores indicateworse outcomes).

Short Form Health Survey (SF-36)

The Short Form Health Survey (SF-36) (McHorney et al., 1993; Ware et al., 1992) is the mostwidely used instrument to measure health-related quality of life. SF-36 consists of 36 items ineight scales: physical functioning (10 items), role limitations caused by physical health problems(4 items), role limitations caused by emotional problems (3 items), social functioning (2 items),emotional well-being (5 items), energy and fatigue (4 items), pain (2 items), and general healthperceptions (5 items). These eight scales can be aggregated into two summary measures: thephysical component score (PCS) and the mental component score (MCS) (Ware, Kosinski, &Keller, 1994). An additional single item assesses change in perceived health. Each scale isdirectly transformed into a 0–100 scale on the assumption that each question carries equalweight. Lower scores indicate more disability.

Multidimensional Scale of Perceived Social Support (MSPSS)

With 4 items for each subscale, the self-administered MSPSS is comprised of 12 items,designed to measure perceptions of social support from family members, friends and significantothers (Zimet, Dahlem, Zimet, & Farley, 1988; Zimet, Powell, Farley, Werkman, & Berkoff,1990). Items are rated on a 7-point Likert Scale (1 = very strongly disagree; 7 = very stronglyagree), with higher scores indicating greater levels of perceived support. Confirmatory factoranalysis has consistently reported a 3-factor solution: family (MSPSS-FA), friends (MSPSS-FR)and significant others (MSPSS-SO). Internal consistency of the Chinese version is good(Tonsing, Zimet, & Tse, 2012; Zhou et al., 2015).

Mindful Attention Awareness Scale (MAAS)

The MAAS is designed to assess open or receptive awareness of and attention to what istaking place in the present. The scale shows strong psychometric properties and has beenvalidated with college students, community samples, cancer patients (Carlson & Brown,2005), and among Chinese populations (Black, Sussman, Johnson, & Milam, 2012).Correlational, quasi-experimental, and laboratory studies have shown that the MAAS taps intoa unique quality of consciousness that is related to, and predictive of, a variety of self-regulationand well-being constructs. The self-report measures are based on 15 items, and the total score isa mean of the 15 items. Higher scores reflect higher levels of dispositional mindfulness. (Brown& Ryan, 2003)

Pittsburgh Sleep Quality Index (PSQI)

PSQI was used for subjective sleep assessment, which included multiple sleep-related vari-ables over the preceding month, using Likert and open-ended response formats (Spira et al.,

6 MA ET AL.

2012). The PSQI yields seven component scores: subjective sleep quality, sleep latency, sleepduration, habitual sleep efficiency, sleep disturbances, sleep medication, and daytime dysfunc-tion. Component scores range from 0 to 3 and are summed to obtain a global score, whichranges from 0 to 21. Higher scores suggest greater sleep disturbance (Buysse, Reynolds, Monk,Berman, & Kupfer, 1989).

ECG Log and Self-Evaluated Sleep Questions

A brief ECG log and self-evaluated sleep questionnaire was used to record the time when theECG recording with the ambulatory device (see 2.5 for description) started, and times thatparticipants went to bed at night and rose in the morning. Participants were asked to recall howmany times they woke up during the night, and evaluate overall sleep quality.

Data Collection

Subjective sleep quality was assessed by Pittsburgh Sleep Quality Index (PSQI) and a brief sleeplog was used to record sleep latency and duration for the studied night. Objective sleep qualitywas assessed by ECG-based cardiopulmonary coupling (CPC) analysis. ECG recordings werecollected by an FDA approved ambulatory electrocardiogram monitor (Dynadx Co.) with acomputer-based data acquisition system. All subjects were monitored at home for 24 hr beforeand after 10 weeks of Tai Chi training. The 24-hr ECG data was analyzed for heart ratevariability, which will be reported elsewhere. Only the ECG recordings during sleep at nightwere extracted for sleep analysis in this study. Sampling rate of ECG monitoring was 200 Hz.All ECG recordings were carefully checked with noise level, artifacts, R peak detection andectopic beats. Data was discarded if quality level was below 90%.

Cardiopulmonary Coupling Analysis of the ECG Recordings

Cardiopulmonary coupling (CPC) analysis is based on mathematical analysis of thecoupling between heart rate variability and the respiratory modulation of QRS waveformon a beat-to-beat basis. Major physiological sleep states derived from CPC analysis includestable sleep (indicated by high-frequency coupling, or HFC), unstable sleep (indicated bylow frequency coupling, or LFC), and REM or wakeful states (indicated by very low-frequency coupling, or VLFC) (Thomas et al., 2005). Stable sleep is associated with beinghealthy, and impaired stable sleep (or increased unstable sleep) is associated with patho-logical conditions (Thomas et al., 2005).

Statistical Analyses

SPSS 19.0 (IBM SPSS Statistics) was used for statistical analyses in this study. Descriptivestatistics were presented as mean ± standard deviation for continuous data, and number (percen-tage) for categorical data. Comparisons of pre- and postmeasurements of continuous variableswere assessed by the nonparametric Wilcoxon Signed Ranks Tests for continuous variables andFisher Exact Tests for categorical variables, in view of the small sample size.

Spearman’s sign-rank correlation tests for nonparametric data were performed to exam-ine the correlations between changes in subjective sleep assessments using PSQI scoresand objective sleep measurements using HFC, LFC, and VLFC, pre and post Tai Chi

TAI CHI FOR SLEEP IN MDD PATIENTS 7

training, and changes in depression and functioning (HAMD-17 scores, BDI, CGI-S, CGI-I, SF-36 scores, and Perceived Stress Mindfulness awareness). A p value < 0.05 wasconsidered statistically significant.

RESULTS

Subjects and Demographics

Nineteen depressed patients were screened for this study, and among them, 16 (6 males and 10females) adult Chinese patients (ages 54.5 ± 11.26 years, range 28–65 years, mean BMI = 23.1 ±2.9 kg/m2) were enrolled, and 3 patients were excluded due to low levels of depressionsymptoms (HAM-D-17 score < 14). All participants completed the 10-week Tai Chi training.All of them were able to perform the 24 basic movements after the training. No adverse eventswere reported. It should be noted that over the course of the study, many participants filled intheir exercise logs at the last minute before they were collected at the beginning of each Tai Chiclass. Our research team questioned the accuracy of such reporting and decided not to analyzedata from the exercise logs.

Subjective Scales and Questionnaire

After the 10-week Tai Chi intervention, both depression outcomes and sleep quality improvedsignificantly (Table 1). For subjective sleep assessments, improved PSQI scores (9.6 ± 3.3 to 6.6± 5.2, p = 0.016) were reported. For depression outcomes, patients had decreases in HAM-D-17(20.1 ± 3.7 to 7.8 ± 5.9, p < 0.001) and BDI scores (22.3 ± 9.1 to 11.1 ± 10.6, p = 0.006).Significant improvements were also found in other clinical measurements including MAAS (3.3± 1.0 to 4.4 ± 1.0, p = 0.006), CGI-S (4.4 ± 0.8 to 2.3 ± 1.3, p = 0.001), CGI-I (4.0 ± 0 to 1.9 ±1.0, p = 0.001), MSPSS-FA (4.6 ± 1.6 to 5.5 ± 1.6, p = 0.018), and SF-36-PCS (41.4 ± 9.3 to47.0 ± 9.8, p = 0.044).

CPC-Based Sleep Assessment

Figure 1 illustrates an improvement of objective sleep stability in a depressed patient after TaiChi training. Changes in objective sleep indices showed that after the 10 weeks of Tai Chiintervention, stable sleep percentage (HPC) increased significantly (31.5 ± 18.7 to 46.3 ± 16.9, p= 0.016), and unstable sleep (LFC) percentage (30.0 ± 22.6 to 14.8 ± 16.6, p = 0.003) decreasedsignificantly. Stable sleep onset latency (75.7 ± 100.6 to 20.9 ± 18.0, p = 0.014) also decreased.No significant change was found in wake or REM sleep percentage (VLFC; 22.1 ± 9.8 to 22.7 ±7.7, p = 0.365) pre and post Tai Chi training (Table 2). Individual changes of outcomes at week 0and week 10 of Tai Chi intervention are shown in Figure 2.

Correlations of the Changes in Depression Outcomes and Sleep Quality

Significant correlations were found between the changes in subjective sleep assessments anddepression outcomes pre and post Tai Chi training (Table 3): ΔPSQI and ΔHAM-D-17 (r = 0.6,p = 0.014), ΔPSQI and ΔBDI (r = 0.62, p = 0.010), ΔPSQI and ΔCGI-S (r = 0.618, p = 0.011),and ΔPSQI and ΔCGI-I (r = 0.569, p = 0.021). However, there were no significant correlations

8 MA ET AL.

between changes in objective sleep measurements (stable sleep onset latency [ΔHFC latency]),stable sleep percentage (ΔHFC), unstable sleep percentage (ΔLFC)), and changes in depressionoutcomes pre and post Tai Chi training. In addition, the correlations between changes insubjective sleep assessments (ΔPSQI) and changes in objective sleep measurements (ΔHFC,ΔLFC) were low and not significant.

DISCUSSION

In this pilot study, our findings suggest that 10-week Tai Chi training may improve sleep qualityin patients with depression, based on patients’ subjective assessments using the PSQI, and theobjective sleep measurements using CPC analysis of continuous sleep ECG data measured withan ambulatory electrocardiogram monitor.

TABLE 1Subjective Assessments and Depression Outcomes (N = 16)

Week 0 Week 10 Z p†

Clinical outcomesHAM-D-17 20.1 ± 3.7 7.8 ± 5.9 –3.520 < 0.001**BDI 22.3 ± 9.1 11.1 ± 10.6 –2.743 0.006**MAAS 3.3 ± 1.0 4.4 ± 1.0 –2.726 0.006**

CGICGI-S 4.4 ± 0.8 2.3 ± 1.3 –3.451 0.001**CGI-I 4.0 ± 0 1.9 ± 1.0 –3.464 0.001**

MSPSSMSPSS-SO 4.0 ± 1.9 5.2 ± 1.6 –1.791 0.073MSPSS-FA 4.6 ± 1.6 5.5 ± 1.6 –2.363 0.018*MSPSS-FR 4.0 ± 1.9 5.0 ± 1.5 –1.228 0.220

SF-36Summary score PCS 41.4 ± 9.3 47.0 ± 9.8 –2.017 0.044*Summary score MCS 36.4 ± 10.2 45.1 ± 10.1 –1.931 0.053

PSQIDuration of sleep 1.6 ± 1 1.3 ± 1.2 –1.897 0.058Sleep disturbances 1.6 ± 0.5 1.2 ± 0.7 –2.111 0.035*Sleep latency 1.8 ± 1 1.1 ± 1.1 –2.484 0.013*Daytime dysfunction 1.9 ± 1 1 ± 1.3 –2.507 0.012*Sleep efficiency 0.8 ± 0.9 0.9 ± 1.2 –0.577 0.564Need medicine to sleep 0 0 − −Overall sleep quality 1.9 ± 0.7 1.1 ± 1 –2.437 0.015*Total PSQI score 9.6 ± 3.3 6.6 ± 5.2 –2.414 0.016*

Values reported are mean ± standard deviation.†Wilcoxon Signed Rank test. *p < 0.05, **p < 0.01Abbreviations: HAM-D-17, Hamilton Rating Scale for Depression; BDI, Beck Depression Inventory; MAAS,

Mindful Attention Awareness Scale; CGI-S, Clinical Global Impressions, Severity scale; CGI-I, Clinical GlobalImpressions, Improvement scale; MSPSS, Multidimensional Scale of Perceived Social Support (SO, significant other;FA, family; FR, friends); SF-36, Short Form Health Survey (PCS, Physical Component Summary; MCS, MentalComponent Summary); PSQI, Pittsburgh Sleep Quality Index.

TAI CHI FOR SLEEP IN MDD PATIENTS 9

Prior studies have reported on the effects of Tai Chi on the treatment of depression (Cho, 2008;Chou et al., 2004; Lavretsky et al., 2011; Yeung et al., 2012). Some previous studies have investigatedthe effects of Tai Chi on insomnia associated with depression. Field and colleagues (Field et al., 2013)reported that Tai Chi combined with yoga reduced prenatal depression, anxiety, and sleep disturbances,while Larkey and colleagues (Larkey et al., 2015) showed that Tai Chi improved fatigue, but notdepression or sleep in breast cancer survivors. This current study, using subjective and objective sleepmeasurements, has added to these small number of studies to support the potential beneficial effects ofTai Chi on insomnia associated with major depressive disorder.

Currently, pharmacological and cognitive behavioral therapy (CBT) are the mainstaytherapies (Ma et al., 2015) for treating insomnia in depressed patients. Due to potentialside effects and the risks of dependence on hypnotics, as well as the shortage of well-trainedtherapists to deliver cognitive behavioral therapy for insomnia (CBT-I) (Deak & Winkelman,2012), sleep disturbances continue to be a common and distressing symptom in patients withdepression. Our findings support that Tai Chi may be a viable option for this population asour results showed that patients who were not Tai Chi practitioners were able to perform the

FIGURE 1 ECG-derived sleep spectrograms from an MDD patient. Left panel: baseline sleep spectrogram; HFC= 12%, LFC = 55%, VLFC = 32%. Right panel: week 10 sleep spectrogram; HFC = 53%, LFC = 24%, VLFC =23%. Sleep quality significantly improved, as shown by decreased stable sleep latency, increased stable sleeppercentage, and decreased unstable sleep percentage.

TABLE 2.CPC-Based Sleep Quality Compared Before and After Tai Chi Training (N = 16)

CPC-based sleep measures Week 0 Week 10 Z p†

Total sleep time in bed (hour) 7.5 ± 1.2 7.2 ± 1.1 –0.804 0.422HFC latency (Stable sleep onset latency, min) 75.7 ± 100.6 20.9 ± 18.0 –2.445 0.014*HFC (stable sleep, %) 31.5 ± 18.7 46.3 ± 16.9 –2.418 0.016*LFC (unstable sleep, %) 45.3 ± 20.1 30.6 ± 16.5 –3.001 0.003**VLFC (wake/REM, %) 22.1 ± 9.8 22.7 ± 7.7 –0.906 0.365

Values reported are mean ± standard deviation.†Wilcoxon Signed Rank test. *p < 0.05, **p < 0.01.Abbreviations: HFC, high-frequency coupling; LFC, low-frequency coupling; VLFC, very low-frequency coupling.

10 MA ET AL.

24 basic movements after 10 weeks of training. Unlike CBT-I, Tai Chi can be used as a self-management approach for patients with depression to avoid the side effects of antidepressantand hypnotic medications. Tai Chi was shown to be safe (Wayne, Berkowitz, Litrownik,Buring, & Yeh, 2014), even for the elderly and physically frail individuals (Adler & Roberts,2006; Gillespie et al., 2012). In addition, Tai Chi is relatively cost-effective, and is reportedto produce high adherence and enjoyment as an intervention in clinical and community-based studies (Sun & Buys, 2015; Tousignant et al., 2014).

Current objective assessments of sleep quality rely primarily on PSG, while the limitations ofPSG have been noted. As a complementary automated technique, cardiopulmonary couplinganalysis has been proposed to quantify sleep stability by using single-lead continuous ECG togenerate sleep spectrograms (Thomas et al., 2005). In this study, impaired sleep quality,characterized by delayed onset of stable sleep and sleep fragmentation, as well as increasedwake or REM and unstable sleep are in line with excessive wakefulness-promoting influencesfound in depressed patients (Yang et al., 2011).

Sleep is a complex physiological process that involves the function of every system atdifferent levels (Bianchi & Thomas, 2013). CPC incorporates the respiration coupling conceptinto the analysis to enable the filtering of power spectra due to nonrespiratory induced HR

FIGURE 2 Change of outcomes at week 0 and week 10 of Tai Chi intervention. (a) HAM-D-17 depressionscore; (b) PSQI total score; (c) Stable sleep percentage, indicated by high-frequency coupling; (d) Unstable sleeppercentage, indicated by low-frequency coupling.

TAI CHI FOR SLEEP IN MDD PATIENTS 11

TABLE

3.Spe

arman

’sCorrelatio

nsforDep

ressionOutco

mes

andSleep

Mea

suremen

tsPre

andPos

tTai

Chi

Interven

tion

Gender

△HAMD

△BDI

△MAAS

△CGI-S

△CGI-I

△SF-36-PCS

△SF-36-M

CS

△PSQ

I△H

FClatency

△HFC

HAMD

r–0.619

*Sig.

0.011

△BDI

r–0.477

0.75

9**

Sig.

0.06

20.00

1△M

AAS

r0.61

7*–0.514

*–0.686

**Sig.

0.011

0.04

20.00

3△C

GI-S

r–0.602

*0.92

5**

0.90

8**

–0.609

*Sig.

0.01

40.00

00.00

00.01

2△C

GI-I

r–0.507

*0.92

5**

0.78

1**

–0.571

*0.897*

*Sig.

0.04

50.00

00.00

00.02

10.000

△SF-36-PCS

r0.22

4–0.624

**–0.560

*0.51

5*–0.629**

–0.726**

Sig.

0.40

40.01

00.02

40.04

10.009

0.001

△SF-36-MCS

r0.53

2*–0.742

**–0.800

**0.611*

–0.787**

–0.783**

0.409

Sig.

0.03

40.00

10.00

00.01

20.000

0.000

0.116

△PSQI

r–0.408

0.60

0*0.62

0*–0.328

0.618*

0.569*

–0.331

–0.641**

Sig.

0.117

0.01

40.01

00.21

40.011

0.021

0.211

0.007

△HFClatency

r–0.112

–0.086

–0.017

–0.114

0.060

–0.176

0.283

–0.013

–0.056

Sig.

0.67

90.75

20.95

00.67

30.824

0.515

0.288

0.961

0.836

△HFC

r0.26

7–0.039

–0.177

0.25

2–0.198

–0.069

0.041

0.227

–0.234

Sig.

0.31

80.88

70.511

0.34

70.463

0.799

0.879

0.398

0.383

0.18

8△L

FC

r–0.042

0.114

0.20

1–0.268

0.241

0.147

–0.046

–0.391

0.205

0.54

1*–0.782**

Sig.

0.87

70.67

40.45

50.31

50.369

0.586

0.867

0.134

0.447

0.03

00.000

Note.

*p<0.05,**p<0.01.

Note.

Abbreviations:HAM-D

-17,

Ham

ilton

RatingScale

forDepression;

BDI,BeckDepressionInventory;

MAAS,Mindful

AttentionAwarenessScale;CGI-S,

Clin

ical

GlobalIm

pressions,

Severity

scale;

CGI-I,Clin

ical

GlobalIm

pressions,

Improvem

entscale;

SF-36,

ShortForm

Health

Survey(PCS,PhysicalCom

ponent

Sum

mary;

MCS,MentalCom

ponent

Sum

mary);PSQI,Pittsburgh

Sleep

QualityIndex;

HFC,high-frequency

couplin

g;LFC,low-frequency

couplin

g.

12

changes, and to enhance the potential diagnostic utility in clinical applications (Ma & Sun, 2013;Stein & Pu, 2012). Compared to conventional analyses using heart rate variability (Ma &McCraty, 2016), our results based on CPC analyses of ECG data integrated the interlinkedphysiological processes including autonomic, respiratory, and electrocortical functions, in addi-tion to brain activities. To our knowledge, this is the first study to have examined the impact ofTai Chi practice on depression and sleep disturbances with ECG-based portable objectiveoutcome indicators. CPC-based sleep spectrograms may complement EEG-based techniquesfor the assessment of sleep stability and quality, and provide low-cost, clinically useful insightinto sleep quality in depressed patients, and possibly into the effects of interventions.

In this study, the improvement of patients’ depression and functioning correlated well withtheir subjective sleep assessment after Tai Chi training, but not with their objective sleepmeasurements using ECG-based CPC analysis. In addition, the correlations between patients’subjective and objective assessments of sleep improvement were low, which have beendescribed as sleep discrepancy in previous studies (Kay, Buysse, Germain, Hall, & Monk,2015; Kay, Dzierzewski, Rowe, & McCrae, 2013; Williams, Kay, Rowe, & McCrae, 2013).The exact reason for such discrepancy is unclear. It is possible that the participants had twodistinctive comorbid conditions, depression and insomnia, and the improvement in one conditionmay not necessarily lead to improvement in the other one. Another possible explanation is thatthe subjective assessment of sleep disturbance may not reflect a person’s true insomnia; rather itreflects a person’s distress as a result of mood symptoms. Given that sleep disturbances arehighly prevalent, portend poorer treatment outcomes, and increase the risk for relapse indepressed populations, understanding the links between sleep quality and depression treatmentoutcome is critical (Troxel et al., 2012).

Limitations

We would like to acknowledge the following limitations of this study. First, this is a pilot studywith a small sample size and no control group in which all subjects received the intervention.While we found significant improvements in subjective and objective sleep measurements, theabsence of a control group obviates the possibility of stating definite conclusions regarding theeffectiveness of Tai Chi for treating insomnia among depressed patients. It is possible thatimprovement of sleep quality was due to the passage of time. Some patients might have beenmotivated to join the study when they were symptomatic and presented with both depression andinsomnia. These symptoms may fluctuate and decrease over time. Second, it is unclear ifpatients’ improvement in the intervention group was a result of Tai Chi or of the socialinteraction from participating in the study, the establishment of a new structure in life, or thestimulation from and engagement in a new commitment to learn and practice Tai Chi. Futureattention-controlled and mechanistic studies might further investigate the differential impacts ofTai Chi and social interaction. Another limitation is the issue of generalizability. As patients inthis study were predominantly recent Chinese immigrants, we cannot be sure whether theseresults would generalize to other populations as there may be cultural beliefs among ChineseAmericans which may lead that population to have favorable expectations about the effects ofTai Chi on sleep. Further studies will be needed to examine if Tai Chi is effective for treatinginsomnia in the mainstream population and in other ethnic minority groups.

TAI CHI FOR SLEEP IN MDD PATIENTS 13

CONCLUSION

Tai Chi training resulted in improvement in sleep quality and mood symptoms among patientswith depression. Our results showed significant correlations between the changes in subjectivesleep measurements and improvement in depression symptoms. The correlations betweenchanges in depression symptoms and changes in objective sleep measurements were low andnot significant.

FUNDING

This study is partially supported by a research grant from South Cove Community HealthCenter.

ORCID

Yan Ma http://orcid.org/0000-0003-1173-2920Albert Yeung http://orcid.org/0000-0001-8018-998X

REFERENCES

Adler, P. A., & Roberts, B. L. (2006). The use of Tai Chi to improve health in older adults.Orthopedic Nursing, 25, 122–126.Agargun, M. Y., Kara, H., & Solmaz, M. (1997). Sleep disturbances and suicidal behavior in patients with major

depression. Journal of Clinical Psychiatry, 58, 249–251.Alberti, S., Chiesa, A., Andrisano, C., & Serretti, A. (2015). Insomnia and somnolence associated with second-generation

antidepressants during the treatment of major depression: A meta-analysis. Journal of Clinical Psychopharmacology,35, 296–303.

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). Washington,DC: Author.Please follow the model of these first four notes. Put them in APA style and alphabetical order.

Bakish, D. (2001). New standard of depression treatment: Remission and full recovery [Suppl]. Journal of ClinicalPsychiatry, 62(26), 5–9.

Beck, A. T., Steer, R. A., Ball, R., & Ranieri, W. (1996). Comparison of Beck Depression Inventories -IA and -II inpsychiatric outpatients. Journal of Personality Assessment, 67, 588–597.

Beck, A. T., Ward, C. H., Mendelson, M., Mock, J., & Erbaugh, J. (1961). An inventory for measuring depression.Archives of General Psychiatry, 4, 561–571.

Bianchi, M. T., & Thomas, R. J. (2013). Technical advances in the characterization of the complexity of sleep and sleepdisorders. Progress in Neuro-Psychopharmacology & Biological Psychiatry, 45, 277–286.

Black, D. S., Sussman, S., Johnson, C. A., & Milam, J. (2012). Psychometric assessment of the Mindful AttentionAwareness Scale (MAAS) among Chinese adolescents. Assessment, 19, 42–52.

Brown, K. W., & Ryan, R. M. (2003). The benefits of being present: Mindfulness and its role in psychological well-being. Journal of Personality and Social Psychology, 84, 822–848.

Buysse, D. J., Reynolds, C. F., 3rd, Monk, T. H., Berman, S. R., & Kupfer, D. J. (1989). The Pittsburgh Sleep QualityIndex: A new instrument for psychiatric practice and research. Psychiatry Research, 28, 193–213.

Carlson, L. E., & Brown, K. W. (2005). Validation of the Mindful Attention Awareness Scale in a cancer population.Journal of Psychosomatic Research, 58, 29–33.

Carney, C. E., Segal, Z. V., Edinger, J. D., & Krystal, A. D. (2007). A comparison of rates of residual insomnia symptomsfollowing pharmacotherapy or cognitive-behavioral therapy for major depressive disorder. Journal of ClinicalPsychiatry, 68, 254–260.

14 MA ET AL.

Cheng P, Goldschmied J, Casement M, Kim HS, Hoffmann R, Armitage R, & Deldin, P. (2015). Reduction in deltaactivity predicted improved negative affect in Major Depressive Disorder. Psychiatry Research, 228, 715–718.

Cho, K. L. (2008). Effect of Tai Chi on depressive symptoms amongst Chinese older patients with major depression: Therole of social support. Medicine and Sport Science, 52, 146–154.

Chou, K. L., Lee, P. W., Yu, E. C., Macfarlane, D., Cheng, Y. H., Chan, S. S., & Chi, I. (2004). Effect of Tai Chi ondepressive symptoms amongst Chinese older patients with depressive disorders: A randomized clinical trial.International Journal of Geriatric Psychiatry, 19, 1105–1107.

Clarke, T. C., Black, L. I., Stussman, B. J., Barnes, P. M., & Nahin, R. L. (2015). Trends in the use of complementaryhealth approaches among adults: United States, 2002-2012. National Health Statistics Reports, 1–16.

Cole, M. G., & Dendukuri, N. (2003). Risk factors for depression among elderly community subjects: A systematicreview and meta-analysis. American Journal of Psychiatry, 160, 1147–1156.

Deak, M. C., & Winkelman, J. W. (2012). Insomnia. Neurologic Clinics, 30, 1045–1066.Don Richardson, J., Cyr, K. S., Nelson, C., Elhai, J. D., & Sareen, J. (2014). Sleep disturbances and suicidal ideation in a

sample of treatment-seeking Canadian Forces members and veterans. Psychiatry Research, 218, 118–123.Emslie, G. J., Rush, A. J., Weinberg, W. A., Rintelmann, J. W., & Roffwarg, H. P. (1990). Children with major depression

show reduced rapid eye movement latencies. Archives of General Psychiatry, 47, 119–124.Field, T., Diego, M., Delgado, J., & Medina, L. (2013). Tai chi/yoga reduces prenatal depression, anxiety and sleep

disturbances. Complementary Therapies in Clinical Practice, 19, 6–10.Frame, K. O., & Alexander, I. M. (2013). Mind-body therapies for sleep disturbances in women at midlife. Journal of

Holistic Nursing: Official Journal of the American Holistic Nurses’ Association, 31, 276–284.Geiger-Brown, J. M., Rogers, V. E., Liu, W., Ludeman, E. M., Downton, K. D., & Diaz-Abad, M. (2014). Cognitive

behavioral therapy in persons with comorbid insomnia: A meta-analysis. In Sleep medicine reviews (Vol. 23C, pp. 54–67).Gillespie, L. D., Robertson, M. C., Gillespie, W. J., Sherrington, C., Gates, S., Clemson, L. M., & Lamb, S. E. (2012).

Interventions for preventing falls in older people living in the community. The Cochrane Database of SystematicReviews, (9), CD007146.

Guy, W. (1976). ECDEU assessment manual for psychopharmacology. US Department of Health, Education, andWelfare, Public Health Service, Alcohol, Drug Abuse, and Mental Health Administration, National Institute ofMental Health, Psychopharmacology Research Branch, Division of Extramural Research Programs.

Hasnain, M., Vieweg, W. V., & Hollett, B. (2012). Weight gain and glucose dysregulation with second-generationantipsychotics and antidepressants: A review for primary care physicians. Postgraduate Medicine, 124, 154–167.

Irwin, M. R., Olmstead, R., Carrillo, C., Sadeghi, N., Breen, E. C., Witarama, T., . . . Nicassio, P. (2014). Cognitivebehavioral therapy vs. Tai Chi for Late Life Insomnia and Inflammatory Risk: A Randomized Controlled ComparativeEfficacy Trial. Sleep, 37, 1543–1552.

Kanai, T., Takeuchi, H., Furukawa, T. A., Yoshimura, R., Imaizumi, T., Kitamura, T. . . . Takahashi, K. (2003). Time torecurrence after recovery from major depressive episodes and its predictors. Psychological Medicine, 33, 839–845.

Kay, D. B., Buysse, D. J., Germain, A., Hall, M., & Monk, T. H. (2015). Subjective-objective sleep discrepancy amongolder adults: Associations with insomnia diagnosis and insomnia treatment. Journal of Sleep Research, 24, 32–39.

Kay, D. B., Dzierzewski, J. M., Rowe, M., & McCrae, C. S. (2013). Greater night-to-night variability in sleepdiscrepancy among older adults with a sleep complaint compared to noncomplaining older adults. Behavioral SleepMedicine, 11, 76–90.

Kennedy, S. (2002). Full remission: A return to normal functioning. Journal of Psychiatry & Neuroscience, 27, 233–234.Kikuchi, T., Suzuki, T., Uchida, H., Watanabe, K., & Mimura, M. (2013). Association between antidepressant side effects

and functional impairment in patients with major depressive disorders. Psychiatry Research, 210, 127–133.Komelski, M. F., Miyazaki, Y., & Blieszner, R. (2012). Comparing the health status of U.S. taijiquan and qigong

practitioners to a national survey sample across ages. Journal of Alternative and Complementary Medicine (New York,NY), 18, 281–286.

Kudlow, P. A., Cha, D. S., Lam, R. W., & McIntyre, R. S. (2013). Sleep architecture variation: A mediator of metabolicdisturbance in individuals with major depressive disorder. Sleep Medicine, 14, 943–949.

Laenen, A., Alonso, A., Molenberghs, G., Vangeneugden, T., & Mallinckrodt, C. H. (2009). Using longitudinal data from aclinical trial in depression to assess the reliability of its outcome scales. Journal of Psychiatric Research, 43, 730–738.

Lai, Y. C., Huang, M. C., Chen, H. C., Lu, M. K., Chiu, Y. H., Shen, W. W. . . . Kuo, P. H. (2014). Familiality and clinicaloutcomes of sleep disturbances in major depressive and bipolar disorders. Journal of Psychosomatic Research, 76, 61–67.

TAI CHI FOR SLEEP IN MDD PATIENTS 15

Larkey, L. K., Roe, D. J., Weihs, K. L., Jahnke, R., Lopez, A. M., Rogers, C. E., . . . Guillen-Rodriguez, J. (2015).Randomized controlled trial of Qigong/Tai Chi Easy on cancer-related fatigue in breast cancer survivors. Annals ofBehavioral Medicine: A Publication of the Society of Behavioral Medicine, 49, 165–176.

Lavretsky, H., Alstein, L. L., Olmstead, R. E., Ercoli, L. M., Riparetti-Brown, M., Cyr, N. S., & Irwin, M. R. (2011).Complementary use of tai chi chih augments escitalopram treatment of geriatric depression: A randomized controlledtrial. The American Journal of Geriatric Psychiatry: Official Journal of the American Association for GeriatricPsychiatry, 19, 839–850.

Li, S. X., Lam, S. P., Chan, J. W., Yu, M. W., & Wing, Y. K. (2012). Residual sleep disturbances in patients remitted frommajor depressive disorder: A 4-year naturalistic follow-up study. Sleep, 35, 1153–1161.

Ma, Y., Dong, M., Mita, C., Sun, S., Peng, C. K., & Yang, A. C. (2015). Publication analysis on insomnia: How muchhas been done in the past two decades?. Sleep Medicine, 16, 820–826.

Ma, Y., & McCraty, R. (2016). Heart rate variability in mind-body interventions. In Complementary Therapies inMedicine.

Ma, Y., & Sun, S. (2013). A new approach to sleep study: Does heart tell us a lot? Sleep Medicine, 14, e188–e9.Ma, Y., Sun, S., & Peng, C. K. (2014). Applications of dynamical complexity theory in traditional Chinese medicine.

Frontiers of Medicine, 8, 279–284.Ma, Y., Zhou, K., Fan, J., & Sun, S. (2016). Traditional Chinese medicine: Potential approaches from modern dynamical

complexity theories. Frontiers of Medicine, 10, 28–32.McHorney, C. A., Ware, J. E., Jr., Raczek, A. E., & The, M. O. S. (1993). 36-Item Short-Form Health Survey (SF-36): II.

Psychometric and Clinical Tests of Validity in Measuring Physical and Mental Health Constructs. Medical Care., 31,247–263.

Montgomery, S. A., & Asberg, M. (1979). A new depression scale designed to be sensitive to change. The BritishJournal of Psychiatry: the Journal of Mental Science, 134, 382–389.

Nietzel, M., & Wakefield, J. (1996). American Psychiatric Association Diagnostic and Statistical Manual of MentalDisorders. Contemporary Psychology, 41, 642–651.

O’Brien, E. M., Chelminski, I., Young, D., Dalrymple, K., Hrabosky, J., & Zimmerman, M. (2011). Severe insomnia isassociated with more severe presentation and greater functional deficits in depression. Journal of Psychiatric Research,45, 1101–1105.

Perlis, M. L., Giles, D. E., Buysse, D. J., Tu, X., & Kupfer, D. J. (1997). Self-reported sleep disturbance as a prodromalsymptom in recurrent depression. Journal of Affective Disorders, 42, 209–212.

Pigeon, W. R., May, P. E., Perlis, M. L., Ward, E. A., Lu, N., & Talbot, N. L. (2009). The effect of interpersonalpsychotherapy for depression on insomnia symptoms in a cohort of women with sexual abuse histories. Journal ofTraumatic Stress, 22, 634–638.

Punnoose, A. R., Golub, R. M., & Burke, A. E. (2012). JAMA patient page: Insomnia. Journal of the American MedicalAssociation, 307, 2653.

Soehner, A. M., Kaplan, K. A., & Harvey, A. G. (2014). Prevalence and clinical correlates of co-occurring insomnia andhypersomnia symptoms in depression. Journal of Affective Disorders, 93–97.

Spira, A. P., Beaudreau, S. A., Stone, K. L., Kezirian, E. J., Lui, L. Y., Redline, S., . . . Stewart, A. (2012). Reliability andvalidity of the Pittsburgh Sleep Quality Index and the Epworth Sleepiness Scale in older men. The Journals ofGerontology Series A, Biological Sciences and Medical Sciences, 67, 433–439.

Stein, P. K., & Pu, Y. (2012). Heart rate variability, sleep and sleep disorders. Sleep Medicine Reviews, 16, 47–66.Sun, J., & Buys, N. (2015). Community-Based Mind-Body Meditative Tai Chi Program and Its Effects on Improvement

of Blood Pressure, Weight, Renal Function, Serum Lipoprotein, and Quality of Life in Chinese Adults WithHypertension. The American Journal of Cardiology, 116, 1076–1081.

Sung, S. C., Wisniewski, S. R., Luther, J. F., Trivedi, M. H., & Rush, A. J. (2015). Pre-treatment insomnia as a predictorof single and combination antidepressant outcomes: A CO-MED report. Journal of Affective Disorders, 174, 157–164.

Thomas, R. J., Mietus, J. E., Peng, C. K., Gilmartin, G., Daly, R. W., Goldberger, A. L., & Gottlieb, D. J. (2007).Differentiating obstructive from central and complex sleep apnea using an automated electrocardiogram-based method.Sleep, 30, 1756–1769.

Thomas, R. J., Mietus, J. E., Peng, C. K., & Goldberger, A. L. (2005). An electrocardiogram-based technique to assesscardiopulmonary coupling during sleep. Sleep, 28, 1151–1161.

Tonsing, K., Zimet, G. D., & Tse, S. (2012). Assessing social support among South Asians: The multidimensional scaleof perceived social support. Asian Journal of Psychiatry, 5, 164–168.

16 MA ET AL.

Tousignant, M., Corriveau, H., Kairy, D., Berg, K., Dubois, M. F., Gosselin, S., . . . Danells, C. (2014). Tai Chi-basedexercise program provided via telerehabilitation compared to home visits in a post-stroke population who havereturned home without intensive rehabilitation: Study protocol for a randomized, non-inferiority clinical trial. Trials,15, 42.

Troxel, W. M., Kupfer, D. J., Reynolds, C. F., 3rd, Frank, E., Thase, M. E., Miewald, J. M., & Buysse D. J. (2012).Insomnia and objectively measured sleep disturbances predict treatment outcome in depressed patients treated withpsychotherapy or psychotherapy-pharmacotherapy combinations. Journal of Clinical Psychiatry, 73, 478–485.

Ware, J. E., Kosinski, M., & Keller, S. (1994). SF-36 physical and mental health summary scales: A user’s manual:Health Assessment Lab.

Ware, J. E., Jr., Sherbourne, C. D., & The, M. O. S. (1992). 36-item short-form health survey (SF-36). I. ConceptualFramework and Item Selection. Medical Care., 30, 473–483.

Watanabe, N., Furukawa, T. A., Shimodera, S., Katsuki, F., Fujita, H., Sasaki, M., . . . Perlis, M. L. (2015). Cost-effectiveness of cognitive behavioral therapy for insomnia comorbid with depression: Analysis of a randomizedcontrolled trial. Psychiatry and Clinical Neurosciences, 69, 335–343.

Wayne, P. M., Berkowitz, D. L., Litrownik, D. E., Buring, J. E., & Yeh, G. Y. (2014). What do we really know about thesafety of tai chi?: A systematic review of adverse event reports in randomized trials. Archives of Physical Medicine andRehabilitation, 95, 2470–2483.

Williams, J. M., Kay, D. B., Rowe, M., & McCrae, C. S. (2013). Sleep discrepancy, sleep complaint, and poor sleepamong older adults. The Journals of Gerontology Series B, Psychological Sciences and Social Sciences, 68, 712–720.

Yang, A. C., Yang, C. H., Hong, C. J., Tsai, S. J., Kuo, C. H., Peng, C. K., . . . Thomas, R. J. (2011). Sleep stateinstabilities in major depressive disorder: Detection and quantification with electrocardiogram-based cardiopulmonarycoupling analysis. Psychophysiology, 48, 285–291.

Yeung, A., Lepoutre, V., Wayne, P., Yeh, G., Slipp, L. E., Fava, M., . . . Fricchione, G. L. (2012). Tai chi treatment fordepression in Chinese Americans: A pilot study. American Journal of Physical Medicine & Rehabilitation /Associationof Academic Physiatrists, 91, 863–870.

Zhou, K., Li, H., Wei, X., Yin, J., Liang, P., Zhang, H., . . . Zhuang, G. (2015). Reliability and validity of themultidimensional scale of perceived social support in Chinese mainland patients with methadone maintenancetreatment. Comprehensive Psychiatry, 60, 182–188.

Zimet, G. D., Dahlem, N. W., Zimet, S. G., & Farley, G. K. (1988). The multidimensional scale of perceived socialsupport. Journal of Personality Assessment, 52, 30–41.

Zimet, G. D., Powell, S. S., Farley, G. K., Werkman, S., & Berkoff, K. A. (1990). Psychometric characteristics of theMultidimensional Scale of Perceived Social Support. Journal of Personality Assessment, 55, 610–617.

TAI CHI FOR SLEEP IN MDD PATIENTS 17