Dietary interventions to improve metabolic health in schizophrenia: A systematic literature review of systematic reviews

Aim: This review of systematic reviews synthesised evidence on the impact of dietary interventions on anthropometric and biochemical measures associated with schizophrenia and metabolic syndrome. Secondly, an aim to identify intervention elements associated with greater dietary adherence and behaviour change . Methods: Five databases were searched from 2000 – March 2023. Eligible reviews included adults, majority diagnosed with schizophrenia, dietary intervention components and at least one anthropometric or biochemical outcome related to metabolic syndrome. Two independent reviewers performed article selection, data extraction, and quality assessment. Results: Seven systematic reviews, consisting of 79 unique primary papers were included. No reviews exclusively examined dietary interventions. Nutrition education and counselling administered alongside physical activity were common. All reviews favoured intervention over the control to reduce body weight, body mass index


Introduction
Schizophrenia is a complex and severe mental disorder affecting up to 1 % of individuals globally, causing significant disability to many aspects of life (Stępnicki et al., 2018).Consumers living with schizophrenia experience a significant reduction in life expectancy; twentythree years fewer than the general population (Firth et al., 2019;NSW Ministry of Health, 2021).Growing evidence suggests this gap is attributed to an elevated risk of metabolic syndrome (MetS) and associated morbidities (Firth et al., 2019).
Antipsychotic medications, widely prescribed to alleviate positive disease symptoms, are a primary risk factor for MetS (Liao et al., 2021), though newer treatments tend to have fewer side effects (Haddad and Correll, 2018).Sedentary lifestyles, substance abuse, poor-quality diets, and deleterious social determinants of health, like poverty and stigmatisation, common to these consumers may increase the risk of MetS (NSW Ministry of Health, 2021; Liao et al., 2021;Shim and Compton, 2020;Olker et al., 2016).Consumers may also have a biological predisposition for MetS (Liao et al., 2021).In combination, these factors present significant barriers for consumers with schizophrenia to maintain physical and mental health, and engagement in treatment.
The prevalence of MetS among Australians with schizophrenia is at least double the general population and up to 80 % of inpatient mental health consumers are living with obesity (Tirupati and Chua, 2007;John et al., 2009;Malhi et al., 2010).This is of concern because consumers with schizophrenia and MetS have a higher likelihood of poorer compliance with antipsychotic medication and experience greater severity of psychotic symptoms and illness recurrence rates (Godin et al., 2018;Dibonaventura et al., 2012;Arango et al., 2008).Body weight is also a significant mental burden for consumers which can interfere with their daily activities and quality of life (Mueller-Stierlin et al., 2022).As such, there is an urgent need to investigate and implement efficacious interventions to reduce the risk of MetS in these consumers.
Lifestyle advice is recommended in numerous treatment guidelines for primary prevention of metabolic morbidities (Firth et al., 2019;NSW Ministry of Health, 2021;World Health Organisation, 2018;Stubbs et al., 2018).Individualisation of this advice is important as public health strategies for the general population are not always effective in those with mental illness (Firth et al., 2019).Recent research focused on physical activity has led to explicit recommendations for treatment providers and consumers (Stubbs et al., 2018).Meanwhile, dietary recommendations remain broad, though it is acknowledged that dietary interventions can complement traditional pharmacotherapies (World Health Organisation, 2018).
An array of dietary interventions have been trialled and reviewed in different settings (Teasdale et al., 2017).Successful interventions include whole-of-diet transitions to Mediterranean and ketogenic patterns, calorie restriction, various supplements, and counselling strategies (Teasdale et al., 2017;Burrows et al., 2022a;Rocks et al., 2022;Onaolapo and Onaolapo, 2021;Sethi and Ford, 2022).All have had considerably different intervention elements and effects on metabolic health.Currently, no umbrella summary exists of these interventions, for schizophrenia exclusively, making it challenging for the research to be interpreted in a useful manner for implementation by health professionals.This becomes more intricate due to the prevalence of comorbid conditions among individuals diagnosed with schizophrenia, notably other mental illnesses like depression and bipolar disorder.Recent reviews have explored nutrition and dietary interventions in individuals with serious mental illnesses (Teasdale et al., 2017;Burrows et al., 2022a).However, to our knowledge, there has not been a specific focus solely on schizophrenia diagnoses in these investigations.
Therefore, specific to the population of adults living with schizophrenia and prescribed antipsychotic medications, the objectives of this review were to; (i) synthesise evidence on the impact of dietary interventions on anthropometric and biochemical measures associated with MetS; and (ii) identify intervention elements associated with greater dietary adherence and behaviour change.

Methods
The protocol for this umbrella review of systematic reviews was registered with PROSPERO (Registration CRD42023407926); and the Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) guidelines were observed (Page et al., 2021).The Joanna Briggs Institute Handbook Chapter 9, 'Umbrella Reviews' were utilised to guide the undertaking of this umbrella review of systematic reviews (Aromataris et al., 2024).
A systematic search was conducted in March 2023 utilising relevant medical subject headings and keywords in the five following electronic databases: MEDLINE, Cochrane Database of Systematic Reviews, Psy-cINFO, CINAHL and SCOPUS.A sample completed Medline search strategy has been provided in the Supplementary Material, Table S1.A limit on publication period was set to the year 2000, as investigations into the effect of dietary interventions were seen to emerge around this time.All papers were limited to the English language.Reference lists from the identified articles were hand searched.No grey literature was searched or reported on, owing to the inclusion criteria.
Criteria for inclusion were peer-reviewed published systematic reviews that met the following conditions: (i) followed a systematic protocol and included only intervention studies.Reviews including observational studies were excluded.(ii) the investigated population were adults, over the age of 18 and this age limit was present as an inclusion criterion within each trial.Within each review, at least 70 % of included participants, summated from each trial, had a diagnosis of schizophrenia or schizophrenia spectrum disorder.Where the specific mental illness of participants was not disclosed, trials were excluded from the calculation.For each review, at least 70 % of included participants were reported to be prescribed any type of antipsychotic medication.Where this was not stated, trials were excluded from the calculation.Citations and abstracts of retrieved articles were imported to EndNote and Covidence software (Veritas Health Innovation, 2023; The Endnote Team, 2013).Once duplicates were removed, two reviewers (AM and TR) independently screened by study title and abstract based on the predetermined criteria.Eligible full texts were then independently assessed by the same two reviewers for final inclusion.Consensus was first attempted to resolve any disagreements.If an agreement was not established, a third reviewer (GM or JC) arbitrated.
PRISMA guidelines informed the development of a standardised Excel form for data extraction.This form was utilised by two independent reviewers (AM and TR) to extract relevant data from included reviews.Characteristics of the reviews extracted were, author and year of publication, research aim, search scope, studies retrieved, participant characteristics, setting, intervention characteristics (type, format, provider, length, delivery mode and adjunctive interventions), comparison groups and funding sources.Quantitative data extracted were anthropometric and biochemical measurements associated with metabolic syndrome, including weight changes, BMI and WC, glucose, blood, and cholesterol biomarkers.Finally, adherence to dietary interventions, assessments of behaviour and knowledge changes, dropout rates and results of subgroup analyses were extracted.Disagreements were resolved through consensus between the two reviewers.
Risk of bias for each included review was assessed by two independent reviewers (AM and TR) using the Risk Of Bias In Systematic Reviews (ROBIS) tool (Whiting et al., 2016).ROBIS assesses four domains: study eligibility criteria; identification and selection of studies; data collection and study appraisal; synthesis and findings.An overall assessment of bias in the reviews of the above four domains, in combination with the reviews interpretation of findings and consideration of limitations, led to an overall risk of bias rating as High, Low or Unclear.
Furthermore, overlapping of studies between different reviews was determined by calculating the percentage of primary studies that were included in more than one systematic review using a validated corrected cover area (CCA) method (Pieper et al., 2014).CCA scores could be interpreted as having slight overlap (0-5 %), moderate overlap (6-10 %), high overlap (11-15 %) and very high overlap (>15 %) (Pieper et al., 2014).
A narrative synthesis with vote counting, median and range calculations was utilised.

Results
The searches were conducted on March 8, 2023, and retrieved 3622 potentially eligible articles.After eliminating duplicates and completing title and abstract screening, 126 articles remained for full text screening.One hundred and nineteen articles were excluded through this process.
Reasons for exclusion are documented in the PRISMA flow chart of review selection, Fig. 1.Seven review articles were considered eligible for this systematic review of reviews.
Characteristics of the seven included reviews are summarised in Table 1.Reviews were published between 2011 and 2022.Across the included seven reviews, there was 79 unique trials, with the sum of unique participants, equalling 1688.Mean age of participants was reported in five reviews (Olker et al., 2016;Bradley et al., 2022;McBain et al., 2016;Speyer et al., 2019;Verhaeghe et al., 2011) and ranged from 39 to 54 years.Gender was relatively even across reviews when reported (n = 5) (Olker et al., 2016;Bradley et al., 2022;McBain et al., 2016;Speyer et al., 2019;Verhaeghe et al., 2011).On two occasions where participant BMI was reported, the mean was above a healthy weight range (BMI > 30 kg/m 2 ) (Speyer et al., 2019;Verhaeghe et al., 2011).No further common characteristics of participants were reported.
There were no reviews which focused exclusively on dietary interventions, all were targeting multiple health behaviours.Included dietary components were heterogeneous in nature.Interventions were categorised into the following types: nutrition education, nutrition counselling, psychoeducation with a nutrition component, nutritional cognitive behavioural therapy, manual based nutrition education, calorie restricted diet, incentives for behaviour change, goal setting and   (Olker et al., 2016;Bradley et al., 2022;Gurusamy et al., 2018;McBain et al., 2016;Speyer et al., 2019;Verhaeghe et al., 2011) reported on a mixed format of intervention, including individual, group, or combination configurations.Similarly, various providers of interventions were described (n = 3) (Bradley et al., 2022;Gurusamy et al., 2018;McBain et al., 2016) and four reviews provided no information (Olker et al., 2016;Singh et al., 2018;Speyer et al., 2019;Verhaeghe et al., 2011).All interventions were delivered face to face, with the length ranging from one week to 18 months.
A summary of outcomes for the included reviews are listed in Table 2. Overall, the reviews indicated improvements in anthropometric and biochemical measurements associated with metabolic syndrome, following interventions with dietary components.Five reviews, with 79 unique trials concluded that such interventions that address diet along with other lifestyle factors were effective or relevant for consumers living with schizophrenia that were on antipsychotic medication (Olker et al., 2016;Bradley et al., 2022;Gurusamy et al., 2018;Singh et al., 2018;Verhaeghe et al., 2011).
For the primary outcome of body weight, four reviews conducted a meta-analysis, and all observed a statistically significant (p < 0.05) effect of interventions for weight loss in comparison to the control (Olker et al., 2016;Bradley et al., 2022;Singh et al., 2018;Speyer et al., 2019).Weight reduction ranged from 1.42 to 2.20 kg in three reviews (Olker et al., 2016;Bradley et al., 2022;Speyer et al., 2019).Another review reported a moderate effect size (Singh et al., 2018).One review analysed weight at a long term follow up and the effect was also statistically significant in favour of the intervention (McBain et al., 2016).Weight maintenance was measured in one review, showing an effect size of − 2.05 kg in favour of the intervention (Speyer et al., 2019).
A statistically significant effect was also reflected for BMI, in which the same four reviews performed a meta-analysis (Olker et al., 2016;Bradley et al., 2022;Singh et al., 2018;Speyer et al., 2019).BMI decreased between 0.48 and 1.17 kg/m2 and a moderate effect size was found in one review (Singh et al., 2018).Where a narrative synthesis was employed, BMI and body weight was seen to reduce for participants, at a statistically significant level (Gurusamy et al., 2018;McBain et al., 2016;Verhaeghe et al., 2011).A review (McBain et al., 2016) narratively examined BMI at long term follow up; this effect was favourable to the intervention, but was no longer statistically significant.
For WC, five reviews reported a decrease in the intervention group (Olker et al., 2016;Bradley et al., 2022;Gurusamy et al., 2018;Singh et al., 2018;Speyer et al., 2019).WC reduction ranged from 0.87 cm to 2.10 cm in three reviews (Olker et al., 2016;Bradley et al., 2022;Speyer et al., 2019) that completed a meta-analysis with the fourth analysis finding a large effect size (Singh et al., 2018).WC was narratively synthesised in one review which found a decrease following interventions (Gurusamy et al., 2018).
Subgroup analyses of weight and BMI were conducted in three reviews (Olker et al., 2016;Singh et al., 2018;Speyer et al., 2019).One review (Speyer et al., 2019) found that individual sessions produced greater reductions in BMI compared to group sessions.In contrast, a more recent review (Olker et al., 2016) identified a mixed intervention format (of individual and group sessions) produced a greater effect on weight reduction than only an individual or group format (− 2.63 kg vs. − 2.50 kg vs. − 0.96 kg, respectively).Adjunctive treatment components were also sub-analysed by two separate reviews.Olker et al. (2016) observed that nutrition counselling along with led or supervised physical activity components resulted in larger weight change compared to nutrition counselling with only education or advice on physical activity (− 2.77 kg vs − 0.90 kg, respectively).Whereas Singh et al. (2018) investigated cognitive behavioural therapies and psychoeducation interventions compared to treatment as usual (TAU), concluding greater reductions in weight and BMI with either CBT or psychoeducation interventions.Studies that emphasised prevention of weight gain, as well as interventions that focused on weight loss were both found to be successful approaches.Prevention studies yielded a larger mean weight loss than weight reduction interventions in one analysis (Olker et al., 2016).Another review found no difference between approaches (Speyer et al., 2019).Shorter active treatment periods, being three months ideally, resulted in statistically significant weight and BMI losses compared to TAU in one review (Olker et al., 2016).Similarly analysis by Singh et al. (2018) concluded a short intervention of four months to be the optimal length to assess change in weight and BMI.The same review (Singh et al., 2018) deduced no difference in effectiveness of interventions performed in developed or developing countries, when targeting weight and BMI reduction.Comparably another review (Speyer et al., 2019) found Asian trials to produce more significant reductions of BMI in comparison to the USA and Europe.
Biochemical results were not routinely reported in the reviews.Reports on blood glucose were seen in three reviews (Gurusamy et al., 2018;McBain et al., 2016;Speyer et al., 2019).Through a meta-analysis, Speyer et al. ( 2019) observed a small increase in blood glucose following  Dropout rates for groups were reported in four studies and ranged from 17.2 %-20.6 % for intervention groups and from 11.5 %-18.4 % for control groups (Olker et al., 2016;McBain et al., 2016;Speyer et al., 2019;Verhaeghe et al., 2011).A review (Speyer et al., 2019) examined the difference of dropout rates between groups and reported no statistically significant differences.Only one review (McBain et al., 2016), which included one trial, stated the intervention provider and this was a diabetes trained mental health professional.This review had an 18.8 % drop out rate and 81 % adherence rate to the intervention.Due to insufficient reporting in other reviews, we were unable to explore any links between studies, intervention providers, dropout rates, dietary adherence, and behaviour change.
There were two reviews that outlined changes in dietary behaviours (Bradley et al., 2022;McBain et al., 2016).Bradley et al. (2022) metaanalysed three trials for fruit and vegetable consumption, which increased slightly compared to the control group but was only significant for vegetable consumption.McBain et al. (2016) reported a reduced amount of calorie consumption by participants in the one trial it reviewed, both immediately after the intervention and at long term follow up.
The level of evidence (Table 3) for the reviews were as follows: two reviews were found to be suggestive (class III) (Olker et al., 2016;Singh et al., 2018), two reviews had weak evidence (class IV) (Bradley et al., 2022;Speyer et al., 2019) and one review was deemed non-significant A. Morris et al. (Gurusamy et al., 2018).Two reviews (McBain et al., 2016;Verhaeghe et al., 2011) did not report p-values and thus, could not be graded.The main reasons for the high risk of bias assessments were inappropriate search strategies and selection procedures; and failure to adequately address risks of bias during synthesis and interpretation of results.The synthesis and findings domain has been rated unclear where a narrative synthesis was performed (Gurusamy et al., 2018;McBain et al., 2016;Verhaeghe et al., 2011).Such reviews provided no information regarding heterogeneity or robustness of the findings.Since one review (McBain et al., 2016) solely reported on one trial, the synthesis domain was marked unclear.
Across the seven reviews, there were 79 unique papers.The total number of papers included was 160, and thus there were 81 papers which overlapped.From calculations, the degree of overlap was determined to be 16.4 %, indicating a very high overlap of studies between reviews.

Discussion
This umbrella review of systematic reviews found that dietary and lifestyle interventions may have a small to moderate effect on reducing anthropometric outcomes for consumers living with schizophrenia and prescribed antipsychotic medications.Overall, the evidence suggests that compared to control or TAU groups, dietary interventions reduce body weight, BMI, and WC and can support weight management and maintenance.However, a level of caution should be exercised when interpreting these results, as over half of the reviews had a high risk of bias with non-significant to weak levels of evidence.Additionally, there was a very high overlap of studies between reviews, indicating a high risk of double counting, which may overinflate findings.Mixed findings were observed for the impact of diet on biochemical parameters of MetS.Adherence to interventions and changes to dietary behaviours were missing or poorly reported, though there was some indication of a positive effect.
It is known that antipsychotic medications have a strong propensity to increase appetite and weight, dysregulate glycaemic control and hypertriglyceridemia (Liao et al., 2021).Consumers living with schizophrenia who are on antipsychotic medications are at high risk of weight gain (Marteene et al., 2019).It has been reported that a 10 week typical dosage of some antipsychotics, including clozapine and olanzapine, has led to significant weight gain between 2.10 and 4.45 kg (Dayabandara et al., 2017).However, the range of weight gain was highly dependent on the antipsychotic mechanism of action and individual characteristics of the consumer (Dayabandara et al., 2017).
Two reviews implied that the small to moderate changes in weight, BMI or waist circumference were of limited clinical relevance of findings (Bradley et al., 2022;Speyer et al., 2019).However, it is important to acknowledge that the typical 5-10 % weight loss target applicable to the general population may not be relevant to this consumer population.Previous research has demonstrated reduced symptom severity, increased cognition, and quality of life for consumers living with schizophrenia, even from similar weight changes and dietary interventions (Adamowicz et al., 2020;Aucoin et al., 2020).Achievement of preventative weight gain, maintenance of weight loss or even small decreases via interventions should therefore be recognised as having clinical significance in improving a consumer's metabolic health outcomes (Firth et al., 2019;NSW Ministry of Health, 2021).This is even more relevant for individuals on long-term usage of antipsychotics, where tolerance to medication can lead to increases in required dosage (Samara et al., 2018).In turn, this may stimulate a stronger appetite and desire for food consumption and would put this consumer again, at an even higher risk for weight gain.
Within Australia, 54.8 % of adults diagnosed with schizophrenia additionally met the criteria for MetS (Cooper et al., 2012).Practice guidelines recommend regular metabolic screening of consumers with schizophrenia, MetS and prescribed antipsychotic medications and a whole-of-lifestyle approach (Firth et al., 2019;NSW Ministry of Health, 2021;World Health Organisation, 2018).Only three reviews examined other MetS specific outcomes, with mixed findings observed for glycaemic control, blood pressure and triglycerides (Gurusamy et al., 2018;McBain et al., 2016;Speyer et al., 2019).This could be attributed to the short trial lengths and a small data pool for follow up periods, minimising the opportunity to observe biochemical changes, and a wide individual variation that may not be adequately captured by mean results (Firth et al., 2019;Liao et al., 2021).Routine biochemical reporting could be achieved with little burden to the researcher when studies are performed in combination with healthcare providers.Furthermore, applying more holistic evaluations of health, such as the Canadian Edmonton Obesity Staging System that considers the metabolic, physical, functional and psychological parameters to determine the most optimal lifestyle and weight management treatment could be more appropriate for this group of consumers than examining anthropometric and metabolic outcomes in isolation (Bradley et al., 2022;Sharma and Kushner, 2009).
Adherence to interventions may have impacted the small to moderate metabolic health effects observed across reviews.Adherence was only reported in one review (McBain et al., 2016).While all systematic reviews included interventions with multimodal lifestyle approaches, dietary intervention components were mostly educational in nature, consistent with a recent literature overview of severe mental illnesses (Burrows et al., 2022a).Tailoring to consumer needs, knowledge and levels of functioning are important (Olker et al., 2016;Gurusamy et al., 2018).Consumers living with schizophrenia experience social determinant barriers outside of knowledge and support, which can impact engagement and adherence to positive lifestyle changes (Speyer et al., 2019;Janz and Becker, 1984;Song and Song, 2023).Previous interventions for consumers with severe mental illnesses have addressed social determinants through food provision and developing food purchasing and preparation skills (Onaolapo and Onaolapo, 2021;Samara et al., 2018;Alpaugh et al., 2020;Jean-Baptiste et al., 2007;McCreadie Table 3 Risk of Bias for each review assessed using the ROBIS Tool (Whiting et al., 2016) and stratification of the level of evidence using a classification method (Fusar-Poli andRadua, 2018) (n = 7). et al., 2005;Teasdale et al., 2016;Dash et al., 2022;Golofast and Vales, 2020).These trials resulted in greater weight loss and dietary adherence rates than those compiled by this review (Onaolapo and Onaolapo, 2021;Samara et al., 2018;Alpaugh et al., 2020;Jean-Baptiste et al., 2007;McCreadie et al., 2005;Teasdale et al., 2016;Dash et al., 2022;Golofast and Vales, 2020).To better capture efficacy of interventions, measurements of adherence and engagement as well as consideration of social barriers need to be incorporated into future trials.Some behaviour changes were noted in the reviews.Favourable increases in vegetable servings and a decrease in energy consumed per day were found.However, as this analysis was based on two reviews containing four unique trials out of seventy-nine, its statistical power and applicability may be limited (Bradley et al., 2022;McBain et al., 2016).Future research should explore food knowledge, relationship with food, food security and accessibility, overall diet quality and nutritional adequacy among this consumer group.Having comprehensive information around the factors influencing consumers' decisions to adopt a healthy diet could be combined with anthropometrics and metabolic health assessments, to further strengthen knowledge, and shape current practices.Similarly, cultural awareness and social reinforcement through peer and individualised support are paramount (Olker et al., 2016).These features could plausibly improve intervention effectiveness as they encompass personal, behavioural, and environmental factors not traditionally included in the biomedical approach to care of chronic poor physical health (Farre and Rapley, 2017).

Research and practice implications
According to the Royal Australian and New Zealand College of Psychiatrists (RANZCP) and their clinical guidelines for consumers with schizophrenia, annual blood tests should be conducted as standard care (Galletly et al., 2016).Reporting of blood results, was however, lacking in the trials gathered by the reviews included in this study.Especially as severely mentally ill patients are less likely to seek out medical help, responsibility should fall on the multidisciplinary team to educate, counsel, and follow-up for overall care (DEH et al., 2011).It is also recommended that regular monitoring of anthropometric and associated metabolic parameters as a part of a thorough and unified healthcare, be undertaken (De Hert et al., 2009).
Data shows that fewer than 5 % of consumers expressing a desire to their general practitioner for lifestyle interventions, are accessing them (Firth et al., 2019;Burrows et al., 2022a;Busch et al., 2016).The findings from this review suggest that mental healthcare services provided across various settings should include a dietary aspect that can be routinely offered to this consumer group.Importantly, the diverse results of this review highlight the need for clinicians from all specialties to be aware of growing research into nutrition and mental healthcare.Previous studies have found dietitian led interventions to be more successful at achieving weight loss and improvements in clinical outcomes for mental health (Teasdale et al., 2017).While cost-effectiveness and long-term sustainability of lifestyle and dietary interventions have yet to be explored specifically in this population; investigations in other mental illnesses are suitable in these aspects, especially when measured with patient centred outcomes (Gurusamy et al., 2018;Burrows et al., 2022b;Park et al., 2013;Segal et al., 2020).
The COVID-19 pandemic is anticipated to have further increased MetS prevalence in this group (Barlati et al., 2021).The pandemic brought disruption to all facets of life including health care (Ayittey et al., 2020).Mental health would likely worsen, with this consumer group finding difficulty in leaving their residence and utilising services offered to them (Kozloff et al., 2020).It is notable that all interventions in these reviews were provided face to face and completed prior to the COVID-19 pandemic.Telehealth is feasible for consumers with schizophrenia and should be considered as an emerging area of research to modify lifestyle behaviours (Santesteban-Echarri et al., 2020).It may also enhance accessibility to health care and intervention adherence, such as for medication adherence, as well as to reduce stigma and improve collaboration with consumer support networks, as observed in other mental illnesses (Santesteban-Echarri et al., 2020;Cohen et al., 2020;Young et al., 2021).

Strengths and limitations
The major strength of this review is that it provides a first-of-its-kind overview and synthesis of research specific to consumers with schizophrenia and prescribed antipsychotic medication; an approach different from previous reviews, in which their focus was on mental health illnesses as a whole and not specified solely to schizophrenia (Teasdale et al., 2017;Rocks et al., 2022).This focus was taken due to a clinical need from a major mental health institution in New South Wales, Australia, with the majority of clients (>90 %) being diagnosed with schizophrenia or schizoaffective disorder.Although broadening this review could be more applicable to the greater mental health population, this has previously been done in other reviews (Teasdale et al., 2017;Rocks et al., 2022).As there was a need for this summation by the institution, it was chosen to have a sole interest in schizophrenia/ schizoaffective disorder diagnosis.Secondly, the review encompasses a variety of outcome measurements relevant to metabolic syndrome to ensure the review remains relevant in future contexts.Thirdly, a comprehensive eligibility criterion was employed to maintain integrity and specificity to the research aims and ensure the results were most applicable to the population of interest and their healthcare providers.Finally, alignment with PRISMA guidelines ensured consistent and bestpractice reporting (Page et al., 2021).
Results obtained by this review are limited by some factors, predominantly reflecting heterogeneous reporting of primary reviews.For example, the vast heterogeneity of intervention elements and lack of adherence and behaviour change data meant the secondary aim could not be answered.Data pertained in the reviews of systematic reviews are removed by two levels, posing difficulties in synthesis of intervention characteristics and reliance on adequate reporting and definitions from original review authors.This is an acknowledged limitation of the study type and in this report has limited the categorisation of dietary intervention components.Intervention characteristics were typically broad and non-specific; therefore, it is possible that some interventions were inappropriately grouped by this review.Similarly, many trials grouped their dietary interventions alongside with others such as smoking cessation, and supervised physical activity, which could have had a synergistic effect and could have bolstered the results.Such broad intervention reporting also means that novel approaches, like supplementation and the Mediterranean diet, may have been overlooked and growing areas of research were unable to be identified.
In summary, it was found that dietary and lifestyle interventions for consumers living with schizophrenia and prescribed antipsychotic medications can have small to modest impacts on anthropometric outcomes and support the ongoing provision of diet, exercise, and other lifestyle interventions to improve and prevent further deterioration of metabolic health.Establishing the overall clinical impact of these reductions is difficult due to heterogenous reporting of quantitative data.Furthermore, with the very high degree of overlap in studies, suggestive to non-significant levels of evidence, and high risk of bias for over half of the reviews, the strength of and credibility of this evidence is reduced, and caution must be exercised when interpreting the results.Future interventions should focus on combined assessment of mental health, physical health, and patient centred outcomes to capture the holistic impacts on consumers and efficacy of interventions to prevent metabolic health deterioration.Furthermore, uniform reporting of intervention elements will allow for cross comparison of efficacious elements and synthesis of evidence at higher levels.More high-quality and rigorous reviews are required.Clinicians should continue to advocate for the inclusion of dietary interventions and dietitians within the mental health care system when the chances for updating policy and funding

Fig. 1 .
Fig. 1.Preferred reporting items for systematic reviews and meta-analyses flow chart of included studies.
These cut-off percentages were established to maintain relevance and applicability to this review's objectives.(iii) at least 70 % of interventions had to be focused on diet, as defined by authors of each review.This ensured that extracted conclusions remained valid to collected data and generalisable to the broader dietetics' profession.Dietary interventions carried out alongside or as part of a greater weight loss, physical activity or lifestyle intervention were included.Any reviews based on pharmacological interventions were excluded.
(iv) at least one MetS related outcome was measured.This included body weight, body mass index (BMI), waist circumference (WC), blood pressure, total serum cholesterol, low density lipoprotein (LDL) and high-density lipoprotein (HDL) cholesterol, triglyceride levels, blood glucose, systolic blood pressure, insulin, and glycated haemoglobin levels (HbA1c).No restrictions were placed on the setting, location, gender, or presence of comparison groups.

Table 1
Characteristics of the included systematic reviews (n = 7).
self-monitoring.The most common were nutrition education (n = 54) followed by nutrition counselling (n = 21).See Table2for which interventions were reported in each review.Interventions provided adjunctive to dietary components were physical activity encouragement, supervised exercise, smoking cessation, cognitive behavioural therapy, and basic diabetes education.Majority of the reviews (n = 6)

Table 2
Outcomes of the included systematic reviews (n = 7).
Note: All 95 % Confidence Intervals are comparing intervention versus control.Abbreviations: