Exercise Therapy in Adults With Serious Mental Illness: A Systematic Review and Meta-analysis

Robert Pearsall; Daniel J Smith; Anthony Pelosi; John Geddes

Disclosures

BMC Psychiatry. 2014;14(117)

Methods

Eligibility Criteria

Studies met the following criteria for inclusion in the review:

Adults with schizophrenia or other types of schizophrenia-like psychosis, schizoaffective disorders, and bipolar affective disorder irrespective of the diagnostic criteria used, age, ethnicity and sex.
All patients, adults, clients, in the community or in hospital.
All relevant randomised controlled trials.
Interventions where a primary or secondary aim was to promote exercise or physical activity.

Search Methods, and Study Selection

We searched the following electronic databases: Ovid MEDLINE, Embase, CINAHL, PsycINFO, Biological Abstracts on Ovid, and The Cochrane Library (January 2009, repeated May 2013). The systematic search included hand searching of journals, books, cross-referencing and bulletins (e.g. brief reports/brief statement of facts). The search filter, the Cochrane Highly Sensitive Search Strategy, was used to assist in the identification of randomised trials in MEDLINE.^[38]

The abstracts of studies were examined by RP. Full text of the studies that potentially met the eligibility criteria was obtained. Discrepancies were discussed with co-investigators. We checked articles that met the inclusion criteria for duplication of the same data.

Data Extraction and Analysis

Data was extracted by one author (RP) and checked for accuracy by the second (DS). Data was extracted onto prepared forms to include: participants and setting, location, description of the intervention, type of exercise, study size, methodological issues, risk of bias, results, and general comments. All analyses were conducted using Revman Manager version 5.1. We performed a PRISMA evaluation of our meta-analysis using a standard checklist of 27 items that ensure the quality of a systematic review or meta-analysis.^[39] A summary measure of treatment effect was used as different outcome measures were found. The standardized mean difference (SMD) with 95% confidence intervals was calculated as the difference in means between groups divided by the pooled standard deviation. If no standard deviations were found they were calculated from standard errors, confidence intervals, or t values.^[40] Authors were contacted for missing data if analyses could not be completed. Statistical heterogeneity was assessed by the I² test. The degree of heterogeneity was categorised as the following:^[36] 0% to 40% low level of heterogeneity; 30% to 60% moderate heterogeneity; 50% to 90% substantial heterogeneity; 75% to 100%: considerable heterogeneity. Standard mean differences were based on the random-effects model as this would take into account any differences between studies even if there was no statistically significant heterogeneity.^[40]

Quality Assessment

There is no agreed standardised method to assess the quality of studies in systematic reviews. We adapted the Cochrane Collaboration's tool for assessing the risk of bias.^[40] The following recommended domains were considered: sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, and other sources of bias. Each item was rated according to the level of bias and categorised into either low, high, or unclear. The category unclear indicated unclear or unknown risk of bias.^[40]

1 2 3 4 5 6

Next Section

Table 1. Systematic review – exercise interventions (Description of programmes).
Study	Study outline	Description of intervention
Beebe et al. [41]	Study to assess the effectiveness of an exercise programme on the physical and mental health of individuals with schizophrenia.	Intervention to determine whether an exercise programme can improve the physical and mental health of individuals with schizophrenia.
		12 participants, random allocation to two groups, intervention and control group (randomisation schedule by statistician). 16 week intervention.
		Intervention group – treadmill exercise programme, meeting 3 times/week for 16 weeks. Programme consisted 10 minutes warm-up session, treadmill walking, then 10 minutes of cool-down exercises. Treadmill walking session gradually increased from 5 minutes to 30 minutes per episode for rest of study.
		Control Group – no exercise programme offered to this group, until end of study.
		Attendance – 43% to 91% of total sessions offered, 75% attended half of sessions, 50% attended 2/3.
		Assessments – demographics, weight/height/BMI, 6-Minute Walking Distance (6 MWD), Percentage Body Fat, Positive and Negative Syndrome Scale Scores (PANSS).
Skrinar et al. [45]	Study to assess the impact of an exercise programme in individuals with serious mental illness.	Intervention of patients with mood and psychotic illness. To determine the effect of a 12 week exercise programme on physical and mental health measures.
		12 week programme. 2 groups - intervention group and control group. Patients randomly assigned.
		Intervention group – 4 exercise per week for 12 weeks + one health seminar per week. Aerobic and cardiovascular training and cool-down each session, increasing intensity of exercise. Health seminar topics – healthy eating, weight management, stress relief, spirituality and wellness.
		Control group – normal care, although offered exercise programme end of study + asked to keep record of amount of exercise they will do during control phase.
		Assessments – demographics, weight, height, pulse, blood pressure, blood tests, Symptom Checklist- 90- R, Lehman Quality of Life Questionnaire, Boston University Making Decisions Questionnaire, MOS 36- Item Short-Form Health Survey (SF- 36). Pre and post-intervention assessment.
Acil et al. [47]	Study to determine the of 10 week exercise programme v. control, in patients with schizophrenia.	10 week programme using intervention and control group. Mixed group of inpatients and outpatients. Randomised two groups.
		Intervention group – 10 week, 3 days per week, and 40 min. per intervention session.
		Aerobic exercise programme for first 2 weeks with 25 min. per day. Starts with 10 min. work up then 25 min. aerobic exercise. Finally 5 min. cooling down.
		Control group – no details of control group.
		Baseline assessment – following assessments:- demographic data, heart rate, Scale for Assessment for Negative Symptoms (SANS), Scale for the Assessment of Positive Symptoms (SAPS), Brief Symptom Inventory (BSI), World Health Organization Quality of Life Scale-Turkish Version (WHOQOL-BREF-TR).
Marzolini et al. [46]	Study to assess the effectiveness of an exercise programme for individuals with schizophrenia.	Intervention to determine the efficacy of a group-based exercise programme. Multidisciplinary approach using a resistance and aerobic exercise programme.
		Intervention over 12 weeks. 2 groups, exercise group and usual care (control). Randomised allocation.
		Intervention – exercise twice per week for 12 weeks at local recreation centre + once per week additional aerobic exercise session individually or during home-visit from mental health clinician. Advised to exercise to same pace and duration at home.
		Control Group – "usual care" continued. No other intervention except measurement at baseline and 12 weeks.
		Assessments – demographic history, weight/BMI, waist/hip circumference, resting blood pressure, functional exercise capacity (6-Minute Walking Distance), muscular strength (one repetition maximum test), anthropometric measurements, adherence to exercise tests, Mental Health Inventory (MHI), self designed feedback questionnaire.
Beebe et al. [43]	Study to determine the effect of effect of exercise advice and techniques, motivational interviewing skills, and a walking programme compared with the control comprising a walking group only in people with schizophrenia spectrum disorders.	Randomised controlled trial of an intervention to effect of exercise advice and techniques, motivational interviewing skills, and a walking programme compared with the control, comprising a walking group only - in people with schizophrenia spectrum disorders. 16 week walking programme with intervention and control group. WALC-P versus TAC
		Intervention – WALC-S programme – Walk (discuss walking/information/advice), Address Sensations (discussion about discomforts warming up or cooling down, suggestions to minimise problems), Learn about Exercise (information on exercise benefits and barriers), Cue Exercise (calendars or times to start walking, with reminder calls etc.).
		Control group – TAC group (time and attention control) consisted of 4 weekly, 1 hour (8–9 subjects per group) – focussing on health behaviour, smoking, relaxation, medication adherence.
		All patients attended walking group for 16 weeks. Assessments – demographic data, walking group attendance, walking group persistence and compliance.
Methapatara et al. [44]	Study to examine the effects of a walking programme in individuals with schizophrenia.	Study to compare the effects of a walking programme combined with motivational interviewing in patients with schizophrenia who are overweight or obese.
		Randomised open label, parallel controlled trial of 12 week duration. Patients with schizophrenia with a BMI of 23 kg/m2. Intervention compared with the control in ratio 1:1. Programme started pre-discharge in hospital for 1 week. Advised to increase walking post-discharge.
		Intervention Group – five 1 hour sessions. Individual motivational interviews given at first session focussing on adequate daily walking. Second session involved group education on nutrition, exercise, warming up, cooling down, and start of use of pedometers. Minimum of 3000 steps per day recommended level of walking.
		Fourth session group walking. Individuals encouraged to increase walking at other times. Fifth session – feedback on programme/progress.
		Control Group – received usual care only. No pedometer given.
		Assessments demographics,, body weight, waist circumference, CGI-Severity, MMSE-Thai.
Scheewe et al. [42]	Study to assess the effectiveness of exercise therapy on cardiorespiratory fitness (CRF) in individuals with schizophrenia.	Intervention to study the effect of an exercise intervention on CRF in patients with schizophrenia compared with matched controls. Trial part of the TOPFIT project. Twenty four week programme. Sixty-three patients with schizophrenia and 55 controls randomised. Randomisation computer generated, stratified for gender, recruitment site and BMI. Four centres, community recruitment.
		Intervention – 1:1 allocation of exercise or occupational therapy. Exercise consists of cycling, rowing, hiking and muscle exercises. Two times a week for 1 hour. Occupational therapy consisted of computer operations, drawing, sculpting and painting.
		Control group - exercise or life as usual. Assessments – Scheewe et al., 2012: CRF assessed using a Cardiopulmonary exercise test (CPET) – stepwise incremental protocol to exhaustion on an upright cycle ergometer. Terminated at voluntary exhaustion. Physiological measurements – blood pressure, heart rate, oxygen update, peak minute ventilation. Scheewe et al., 2013: PANSS, MADRS, Camberwell Assessment of need (CAN), body mass index, waist circumference, blood pressure, blood parameters. Scheewe et al., 2013: Global brain volumes, hippocampal volume, and cortical thickness.
Pelham et al. [48]	Study to assess the impact of an exercise programme in individuals with serious mental illness.	Intervention to assess the effect of an exercise programme using cycling as the method of exercise activity compared to a control. 12 week programme. Eight patients with schizophrenia and 2 with bipolar affective disorders. Total 10 patients in programme. Randomisation into two equal groups. Based in psychiatric rehabilitation services.
		Intervention – exercise activity on a cycle ergometer. Four sessions per week for 30 minutes at 65% to 75% of heart rate reserve.
		Control – muscle tone and strengthening exercises, 4 times per week for 30 minutes each session. Short intermittent bouts used within this to keep heart rate below 110 beats per minute.
		Assessments – VO₂ max tests, Becks Depression Inventory, and weight measurements.

Table 2. Systematic review – exercise interventions (Methodological issues/Risk of bias).
Study	Country	Participants & setting	Design	Recruitment	Baseline sample	Number completing trial	Methodological issues/Risk of bias
Beebe et al. [41]	USA	Patients attending outpatient clinic at a Veterans Hospital	RCT	Volunteers	12	10	Participants – eligibility criteria specified - yes
							- standardised diagnostic criteria - Yes
							- non-random recruitment
							- similar baseline groups
							Intervention – details of therapists training not indicated
							- no treatment manual
							- compliance of medication non- checked
							Measurement of Outcome
							- outcome assessors blinded to treatment allocation
							- adequate follow- up period
							Risk of Bias - adequate sequence generation- unclear
							allocation concealment- unclear
							blinding of participants, personnel and outcome assessors - yes
							incomplete data outcome - yes
							free of selective outcome reporting- yes
							free of other sources of bias- yes
							Small sample, majority male, control group awareness of exercise programme.
							Attendance at sessions variable.
Skrinar et al. [45]	USA	Patients from inpatient, partial hospitalisation, outpatient and community centres in area of McLean Hospital.	RCT	Selection/invitation	30	20	Participants – eligibility criteria specified - yes
							- standardised diagnostic criteria -yes
							- non-random recruitment
							- similar baseline groups –
							Intervention – details of therapists training not indicated
							- no treatment manual
							Measurement of Outcome
							- outcome assessors not blinded to treatment allocation
							- adequate follow- up period
							Risk of Bias - adequate sequence generation - unclear
							allocation concealment - unclear
							blinding of participants, personnel and outcome assessors - no
							incomplete data outcome - no
							free of selective outcome reporting - yes
							free of other sources of bias - yes
							Small sample, variable adherence to programme and attendance, inclusion criteria include mood disorders as well.
							Control group may be affected by exercise "influence" of study.
Acil et al. [47]	Turkey	Inpatient and outpatient. Diagnosis of schizophrenia	RCT	Recruitment method not described, 30 outpatients and inpatients.	30	30	Participants – eligibility criteria poorly specified
							- standardised diagnostic criteria- yes
							- non-random recruitment
							- similar baseline group
							Intervention – details of therapists training not indicated
							- no treatment manual
							Measurement of Outcome
							- outcome assessors not blinded
							- adequate follow- up period
							Risk of Bias - adequate sequence generation - no
							allocation concealment - no
							blinding of participants, personnel and outcome assessors - no
							incomplete data outcome - unclear
							free of selective outcome reporting - unclear
							free of other sources of bias - yes
							Non-standardized exercise intervention. No measurement of existing exercise level or patient participation in the programme.
							Number completed trial not given.
Marzolini et al. [46]	Canada	Identified from Community Mental Health Programme. Majority lived supported accommodation.	RCT	Volunteers	13	13	Participants – eligibility criteria specified - yes
							- standardised diagnostic criteria -yes
							- non-random recruitment
							- similar baseline groups –
							Intervention – details of therapists training not indicated
							- no treatment manual
							Measurement of Outcome
							- outcome assessors not blinded to treatment allocation
							- adequate follow- up period
							Risk of Bias - adequate sequence generation - yes
							allocation concealment - yes
							blinding of participants, personnel and outcome assessors - unclear
							incomplete data outcome - no
							free of selective outcome reporting - yes
							free of other sources of bias - yes
							Small sample, participants supported accommodation. Inclusion criteria – patients had to have one or more cardiovascular risk factors.
Beebe et al. [43]	USA	Outpatients	RCT	Volunteers	97	79	Participants – eligibility criteria specified
							- standardised diagnostic criteria
							- non-random recruitment
							- similar baseline groups
							Intervention – details of therapists training indicated
							- treatment manual - yes
							- compliance of medication non- not checked
							Measurement of Outcome
							- outcome assessors not blinded to treatment allocation
							- adequate follow- up period
							Risk of Bias - adequate sequence generation - yes
							allocation concealment - unclear
							blinding of participants, personnel and outcome assessors - unclear
							incomplete data outcome - unclear
							free of selective outcome reporting - yes
							free of other sources of bias - yes
							Well-designed programme. Adequate sample size.
							No assessment of mental health changes - improvement in mental health may have contributed to increase in exercise by itself.
Methapatara et al. [44]	Thailand	Inpatient & Outpatients	RCT	Volunteers	64	64	Participants – eligibility criteria specified
							- standardised diagnostic criteria – not indicated
							- volunteers
							- dissimilar baseline groups –
							younger control population
							Intervention – details of therapists training not indicated
							- treatment manual - no
							- compliance of medication not checked
							Measurement of Outcome
							- outcome assessors blinded to treatment allocation
							- adequate follow- up period
							Risk of Bias - adequate sequence generation - yes
							allocation concealment - yes
							blinding of participants, personnel and outcome assessors - no
							incomplete data outcome - yes
							free of selective outcome reporting - yes
							free of other sources of bias - yes
							Small sample size, no record of daily steps recorded or measurement of change in exercise levels. Compliance with programme unknown (as no recording of pedometer). No assessment of nutritional intake which may have affected outcome. Effect on mental health unknown.
Scheewe et al. [42]	USA	Community	RCT	Volunteers	118	92	Participants – eligibility criteria specified - yes
							- standardised diagnostic criteria – not indicated
							- non-random recruitment
							- dissimilar baseline groups
							Intervention – details of therapists training indicated
							- treatment manual
							Measurement of Outcome
							- outcome assessors not blinded to treatment allocation
							- adequate follow-up period
							Risk of Bias - adequate sequence generation - yes
							allocation concealment - yes
							blinding of participants, personnel and outcome assessors - no
							incomplete data outcome - unclear
							free of selective outcome reporting - yes
							free of other sources of bias - yes
							Well-designed programme. Large sample size. Computer generated randomisation. Study part of larger research trials. No results from mental health assessment. Baseline group differences. Mean weight greater in patients than controls. Motivated group of patients with higher level of fitness at baseline. No follow up of participants to show whether improvement in exercise maintained.
Pelham et al. [48]	USA	Community Rehabilitat-ion	RCT	Volunteers	10	10	Participants – eligibility criteria specified - yes
							- standardised diagnostic criteria – not indicated
							- non-random recruitment
							- baseline groups unclear
							Intervention – details of therapists training indicated - no
							- treatment manual- no
							Measurement of Outcome
							- outcome assessors not blinded to treatment allocation
							- adequate follow-up period
							Risk of Bias - adequate sequence generation - unclear
							allocation concealment - unclear
							blinding of participants, personnel and outcome assessors - no
							incomplete data outcome - unclear
							free of selective outcome reporting - unclear
							free of other sources of bias – unclear
							Small study. Control also using exercise activity. Data not available to include in meta-analysis. Early study 21 years ago. Beneficial effects on exercise fitness and depression scores.

Table 3. Systematic review – exercise interventions (Results/Comments).
Study	Results	Comments/Analysis
Beebe et al. [41]	6-Minute Walking Distance (MWD) – improvement in distance clinically (152.5 mins) compared with the control group (56.7 mins), but not statistically significant. BMI and body fat reduced in intervention group compared with the control group but not statistically significant. Control Group – increased 6- MWD during intervention (5%).	Valuable study, limited by small sample, population type. Control group showed some increase in physical activity (? overlap effect of intervention).
Skrinar et al. [45]	Weight change intervention: control −2.2 v. −1.2 kg (non-significant). Exercise intensity increased in intervention compared with the control + 8 Watts v. −5 Watts (non-significant). Significant improvement in results in general health (p < 0.05) and empowerment (p < 0.01).	Valuable study although some limitations in methodology.
Acil et al. [47]	Exercise programme resulted in decrease in psychiatric symptoms and increase in quality of life. Reduced SANS, SAPS and BSI. Increase in WHOQOL-BREF-TR. No exercise measures change of heart rate. Demographic data not provided.	Valuable intervention study, however limited by small sample size, and lack of standardization of intervention. No details of drop-out rate or measurement of exercise levels in subjects.
Marzolini et al. [46]	Mean age 43 years. Exercise group showed a 27.7 metre (SD ± 22.3 m) increase in 6MWD while control group showed decrease of 28.3 metres (SD ± 26.6 m) (between group difference, p = 0.1). There was significant increase in strength exercise and Mental Health Inventory. Attendance averaged 72% (SD ± 4.4%) with no dropouts.	Valuable study but limited by small sample size. Good adherence to programme.
Beebe et al. [43]	Percentage attendance in WALC-S group 35.2 versus 27.3% after 16 week programme. 33.7 versus 22.9 in TAC group. Greater persistence in weeks in WALC-S versus TAC, and higher minutes walked (76.67 versus 116.89) in WALC-S group, and (61.88 versus 788.83) in TAC group.	Well-designed study indicating the benefit of exercise advice with a motivational programme in addition to exercise intervention.
Methapatara et al. [44]	End of programme, mean body weight decreased significantly compared with the control group by 2.21 kg (p = 0.03).	Valuable study showing the benefit of an exercise programme in overweight or obese patients with schizophrenia.
Scheewe et al. [42]	Patients had higher resting HR, lower pear HR, peak systolic BP, relative VO_{2 peak}, W_peak, RER, minute ventilation, and HR recovery than controls. In conclusion patients had lower CRF levels compared with controls. Exercise therapy increased VO_{2 peak}, and W_peak in patients and controls. VO_{2 peak}, and W_peak decreased in non-exercising patients.	Well-designed study. Results show an increase in CRF in individuals with schizophrenia. However individuals may be more motivated and have greater baseline fitness than many individuals with this illness.
	Trend-level effect on depressive symptoms (p = 0.07). No effect on symptoms of schizophrenia.
	Significantly smaller baseline cerebral (gray) matter, and larger third ventricle volume, thinner cortex. NO change global brain, hippocampal volume, or cortical thickness.
Pelham et al. [48]	Aerobic exercise group showed significant increases in fitness and a reduction in depression scores. Non-aerobic groups did not improve in fitness level or level of depression.	Small study of 10 patients. Conducted 21 years ago. Findings showed the possible positive effects of exercise on levels of fitness and mental health in people with serious mental illness.

Table 4. Comparison: exercise versus standard care.
Outcome or subgroup title	No. of studies (available data)	No. of participants	Statistical method	Effect Size (SWD)
Exercise activity	1	13	SMD (IV, Random, 95% CI)	1.81 [0.44 to 3.18]
BMI	4	151	SMD (IV, Random, 95% CI)	-0.24 [-0.56 to 0.08]
Weight	2	77	SMD (IV, Random, 95% CI)	0.13 [-0.32 to 0.58]
Negative	2	84	SMD (IV, Random, 95% CI)	-0.54 [-1.79 to 0.71]
Symptoms
Positive	2	84	SMD (IV, Random, 95% CI)	-1.66 [-3.78 to 0.45]
Symptoms
Anxiety/Depression	3	94	SMD (IV, Random, 95% CI)	-0.26 [-0.91 to 0.39]
Q of L (Physical)	2	30	SMD (IV, Random, 95% CI)	0.45 [-0.27 to 1.18]
Q of L (Mental)	2	30	SMD (IV, Random, 95% CI)	0.65 [-0.09 to 1.39]