Table 1.

Table of characteristics and within-study results for all studies included in the systematic review

Author (year, country)	Study design	Baseline participants (n)	Mean age (years)	Sex	SMI diagnosis	Intervention	Comparison	Outcome measure(s)	Within-study results
Battaglia et al. (2013) (Italy)(45)	RCT	n 23 IG n 12 CG n 11	IG 36·00, sd 5·00 CG 35·00, sd 4·00	100 % male	Schizophrenia and/or schizoaffective disorder	Soccer training: training twice weekly. Technical–tactical strategies to promote a combination of moderate- and vigorous-intensity activity. Setting: unreported	Instructed not to engage in physical activity.	BMI (kg/m2) Wt QoL	Follow-up 12 weeks. IG = 10; CG = 8 IG – significant within-group decrease in BMI and wt (P = 0·001). CG – significant within-group increase in BMI and weight (P = 0·05). Between groups change in SF-12 PCS and SF-12 MCS in favour of TG (P < 0·001). Adherence: > 80 %
Elmslie et al. (2006) (New Zealand)(47)	RCT	n 60	IG 42, sd 10 CG 42, sd 13	n 49 (81·7 %) female	Bipolar disorder	Carnitine supplementation: individualised diet plan, physical activity 5 days per week. Lifestyle advice provided by a dietitian. Setting: community	Placebo capsules. Diet plan, physical activity and lifestyle advice same as intervention group.	BMI (kg/m2) Wt Waist circumference Mood – Montgomery–Asberg Depression Rating Scale (MADRS).	Follow-up 26 weeks. Participants n 44 No significant difference between groups for change in BMI (P = 0·466), wt (P = 0·381) or waist circumference (P = 0·597). No significant difference between groups for MADRS (P = 0·190). Adherence: unreported.
Frank et al. (2015) (USA)(48)	RCT	n 122 IG n 61. CG n 61.	41·6, sd 9·5 IG 41·8, sd 9·5 CG 41·4, sd 9·7	Male and female	Bipolar disorder	Psychopharmacologic management. The Healthy Lifestyle Program: 15 individual sessions. Psychoeducation, wt loss nutrition, physical activity. Setting: outpatients	Psychopharmacologic management without Healthy Lifestyles Program.	BMI (kg/m2)	Follow-up 14 weeks, Participants n 109 IG mean BMI decrease of 2·3 % (sd 3·8). CG mean BMI decrease of 0·2 % (sd 3·9), significant difference between groups in favour of IG (t = 2·8, P = 0·006). Adherence: unreported.
Henderson et al. (2005) (USA)(50)	RCT	n 37 IG n 19 CG n 18	IG 3·2, sd 10·6 CG 40·7, sd 9·9	IG: 7 (37 %) female CG: 7 (39 %) female	Schizoaffective disorder, schizophrenia	Sibutramine: weekly group/individual sessions focused on nutrition, exercise, behavioural modification and goal setting. Setting: outpatient clinic of an urban mental health centre	Placebo capsules Individual and group sessions identical to intervention group.	BMI ((kg/m2) Wt Waist circumference Body composition Global Assessment Scale, Positive and Negative Syndrome, Scale for the Assessment of Negative Symptoms.	Follow-up 12 weeks Participants n 31 Significant between groups difference in BMI (P < 0·05), wt (P < 0·05) and waist circumference (P < 0·005) in favour of IG. Significant increase in waist:hip ratio in IG (P < 0·05). Adherence: unreported.
Masa-Font et al. (2015) (Spain)(55)	RCT	n 332 IG n 169 CG = 163	46·7	45 % female	Schizophrenia, schizoaffective, bipolar disorder	Educational and physical activity programme: education component. Physical activity, increase daily steps in local streets. Healthy dietary habits, diet diary. Setting: Mental Health Centres, local streets.	TAU, regular psychiatrist check-ups.	BMI (kg/m2) Waist circumference QoL SF-36 Mental health – Clinical Global Impression (CGI)	Follow-up 3 months IG n 169; CG n 163 Significant decrease in BMI in favour of the CG (P = 0·038). No significant differences in waist circumference. Adherence: 42·6 %
Sugawara et al. (2018) (Japan)(56)	RCT	n 265 Group A, n 85 Group B, n 93 Group C, n 87	Group A, 44·0 sd, 10·3 Group B, 47·6 sd, 9·6 Group C, 46·6 sd, 10·9	Group A, 57 % (35) male Group B, 46 % (31) male Group C, 53 % (32) male	Schizophrenia or schizoaffective disorder	Group B psychiatrist wt loss advice: body wt measurements taken at individual sessions with brief advice. Notebook to record progress. Group C nutrition education: same as Group B plus structured nutrition education sessions with dietitians once a month. Daily food records. Setting: outpatient	Group A, TAU for schizophrenia.	BMI (kg/m2) Wt Waist circumference	Follow-up 12 months Participants n 189 Within Group C, there was a significant decrease in wt (p = < 0·00), BMI (p = < 0·001) and waist circumference (P = 0·007 from baseline to follow-up but not in Group A or B. Adherence: unreported.
Urhan, Aksoy and Ayer (2015) (Turkey)(57)	RCT	n 30 IG n 15 CG n 15	IG 38·08 sd 4·79 CG 37·71 sd 6·04	100 % female	Schizophrenia	Wt loss diet therapy: nutrition and physical activity education. Personalised diet recipes. Dietary motivation. Exercise, daily 30 min walking at moderate pace. Food and physical activity diary. Setting inpatient	Same as IG (no diagnosis of Schizophrenia.)	BMI (kg/m2) Wt Waist circumference Body fat (%) Hip circumference Waist:hip ratio	Follow-up 8 weeks IG n 13; CG n 14 There were significant decreases in all anthropometric measurements within both groups (P < 0·05). Between groups CG had greater decreases in BMI, wt, waist circumference, body fat, hip circumference and waist:hip ratio Adherence: unreported.
Weber and Wyne (2006) (USA)(58)	RCT	n 17 IG n 8 CG n 9	Not reported	IG n 5 (62·5 %) female CG n 7 (77·8 %) female	Schizophrenia or schizoaffective disorder	Cognitive/behavioural group intervention: 60 min per week. Role playing, goal setting, problem-solving and discussions on barriers to change. Activities such as walking. Presentations on low-fat diets. Food and activity diaries. Setting: public mental health clinics.	TAU at the clinic. Weighed and measured (4-week intervals).	BMI (kg/m2) Wt Waist:hip ratio.	Follow-up 16 weeks IG n 9; CG n 7 Between groups the IG had greater wt reductions. Within groups no significant differences in wt, waist:hip ratio, or BMI scores. Adherence: unreported.
Whicher (2021) (UK)	RCT	n 47 IG n 24 CG n 23.	IG 43·9 sd 11·0 CG 45·4 sd 10·7	n 23 female	First episode psychosis, schizophrenia and schizoaffective disorder	Liraglutide as an adjuvant to promote wt loss. Setting: Mental health centres and primary care	Placebo	BMI Wt Waist circumference Impact on mental health	Follow-up 6 months IG n 15, CG n 19 Mean change Significant difference in favour of IG for BMI, weight and waist circumference.
Wu et al. (2005) (Taiwan (ROC)(59)	RCT	n 56 IG n 28 CG n 28.	IG 42·2 sd 7·5 CG 39·0 sd 6·7	IG n 17 female (61 %) CG n 14 (56 %) female	Schizophrenia	Diet and Physical Activity Programme: dietary assessment, including fruit and vegetables, sugar-free foods and drinks. Physical activity 3 d per week, walking up and downstairs. Rewards for participation Setting: Inpatients of a veteran’s hospital	No description reported	BMI Wt Body fat (%) Waist circumference Hip circumference Waist:hip ratio	Follow-up 6 months IG n 28, CG n 25 Significant difference in favour of IG for BMI, wt, waist circumference (P < 0·001) and hip circumference (P < 0·05). Adherence: unreported.
Abdel-Baki et al. (2013) (Canada)(44)	Quasi-experimental	n 25	25·9 (sd 3·9)	100 % male	Schizophrenia, schizoaffective disorder, bipolar disorder, other psychotic disorder	Aerobic interval training (AIT) twice weekly. High-intensity exercise training of shorter duration involving running and active recovery walks. Setting: unclear	N/A	BMI (kg/m2) Wt Waist circumference Lean mass (%) Muscle mass (%)	Follow-up 14 weeks Participants n 16 Significant decrease in waist circumference only (P = 0·015).
Centorrino et al. (2006) (USA)(46)	Quasi-experimental	n 22.	40·5 (sd 8·5) (n 17)	n 10 (58 %) female.	Schizophrenia or schizoaffective disorder	TRIAD: twice weekly meetings. Dietary counselling and exercise low-calorie food plan. One session every 4 weeks. Setting: weight management centre on hospital grounds	N/A	BMI (kg/m2) Wt QoL – SF-36/Quality-of-Life Questionnaire (QLS)	Follow-up 24 weeks, participants n 17 Significant decrease in BMI (P = 0·0005) and wt (P = 0008). SF-36 results not reported. QLS no significant differences at follow-up. Adherence: n 6 attended weekly sessions.
Galle et al. (2017) (Italy)(49)	Quasi-experimental	n 153 Interpersonal therapy n 50. Dialectical Behavioural Group n 50. TAU = 53	IT: 33, sd 4·21 (22–56) DBG: 34, sd 3·78 (26–47) CG: 32, sd 5·10 (21–48)	100 % female	Borderline personality disorder	12 months IT: evaluation of interpersonal relationships, 60-min weekly sessions, optional telephone consultation. DB: strategies for behaviour Modifications, optional telephone consultation. Setting: unreported	TAU Medical and psychological evaluation, meeting with bariatric surgeon.	BMI (kg/m2) Wt	Follow-up 12 months, participants n 139 Significant difference in BMI between groups in favour of IG group (P < 0·01). Wt outcomes only reported for subgroups. Adherence: > 65 %
Henderson et al. (2009) (USA)(51)	Quasi-experimental	n 24	Clozapine + ziprasidone 48, sd 7 Olanzapine + ziprasidone 52, sd 8	n 17 (81 %) male	Schizophrenia or schizoaffective disorder.	Ziprasidone: as an adjuvant to promote wt loss. Ziprasidone plus clozapine Ziprasidone plus olanzapine Setting: outpatient clinic, urban community mental health centre	No control	BMI (kg/m2) Wt Waist circumference Hip circumference QoL Positive and negative syndrome scale, Hamilton Depression Rating scale, Scale for the assessment of negative symptoms	Follow-up 6 weeks, participants n 21 No significant differences at follow-up for any outcomes. Adherence: unreported.
Katekaru, Minn and Pobutsky (2015) (USA)(52)	Quasi-experimental	n 47	31–60 (n 35) 61 and older (n 12)	15 (32 %) female	Paranoid schizophrenia, undifferentiated schizophrenia, schizoaffective disorder	Health counselling: repetitive behavioural counselling, motivational interviewing, physical activity and nutrition guidance. Weekly wellness classes. Setting: Community Mental Health Centre	N/A	BMI – (kg/m2)	Follow-up 5 years Participants n 47 Baseline: n 22 participants were obese (47 %), n 20 were overweight (43 %) Year 5: n 21 participants were obese (45 %), n 18 were overweight (38 %) Adherence: unreported.
Klam, McLay and Grabke (2006) (Canada)(53)	Quasi-experimental	n 19	26 to 62, median age = 41.	n 9 female	Schizophrenia or schizoaffective disorder	Personal Empowerment Program: a healthy lifestyle and wellness group, weekly sessions for 2 h. Education on nutrition, stress management and personal development. Exercise, core work and stretching. Community/affordable recreational activities. Food diary Setting: outpatient clinic	N/A	BMI (kg/m2)	Follow-up 38 weeks Participants n 16 Total wt loss was 149·6 pounds. Average BMI 37·55 kg at baseline and 36·84 kg/m2 at follow-up. Decrease in abdominal girth 20¾ inches. Adherence: average 80 %
Kuo et al. (2013) (Taiwan -ROC)(54)	Quasi- experimental	Schizophrenia group n 33	37·8, sd 1·7	n 14 female	Schizophrenia	Wt reduction programme: psychosocial evaluation and behaviour therapy. Healthy eating and calorie reduction. Exercise five times a week at mild–moderate intensity. Diet/exercise diaries Setting a hospital day care unit	N/A	BMI (kg/m2) Wt Waist Waist:hip ratio	Follow-up 10 weeks Participants n 33 Significant decreases from baseline to follow-up in BMI, wt and waist circumference (P < 0·01) and in waist to hip ratio (P < 0·05). Adherence: unreported.
Barnard-Kelly (2022) (UK) (From Whicher, 2021)	Nested qualitative (from RCT Whicher, 2021)	Baseline interview n 16 Follow-up interview n 10	21–64	N/A	Schizophrenia, schizoaffective disorder or first episode psychosis	Liraglutide as an adjuvant to promote wt loss. Setting: Mental health centres and primary care	Semi-structured interviews. Content and thematic analyses.	Key themes • Medication-associated weight gain • Impact of study on quality of life • Study information and support from trial was well received • Practical aspects of clinic attendance • Healthcare professional perspective

RCT, randomised controlled trial; IG, intervention group; CG, comparator group; Wt, weight; QoL, quality of life.