Introduction
For many illnesses, lifestyle interventions are seen as the cornerstone of prevention and first-line treatment, because they can improve key clinical outcomes and are generally associated with low risks for side effects [Reference Galaviz, Weber, Suvada, Gujral, Wei and Merchant1–5].
At the same time, there is an urgent need to expand standard treatment options for people with severe mental illness (SMI). SMI, defined as schizophrenia spectrum disorders, bipolar and major depressive disorders in this work [6], lead to a high symptom burden, low levels of functioning in daily life and have very often unfavourable outcomes due to persistent psychiatric symptoms [Reference Collaborators7]. Many affected individuals do not benefit sufficiently from standard psychological and psychopharmacological approaches; in addition, several psychotropic medications are frequently accompanied by adverse side effects on many health aspects, such as weight-gain and metabolic syndrome [Reference Correll, Detraux, De Lepeleire and De Hert8, Reference Pillinger, Howes, Correll, Leucht, Huhn and Schneider-Thoma9], resulting often in a lower treatment adherence [Reference Dayabandara, Hanwella, Ratnatunga, Seneviratne, Suraweera and de Silva10]. SMI is also associated with significant somatic comorbidity. Life expectancy in patients with SMI is reduced by 10 to 20 years compared to the general population, with cardiopulmonary and cardiovascular diseases being among the leading causes of death [Reference Correll, Solmi, Veronese, Bortolato, Rosson and Santonastaso11–Reference Biazus, Beraldi, Tokeshi, Rotenberg, Dragioti and Carvalho14]. Further, reduced access to healthcare and mental-health-related stigma in healthcare lead to lower rates of screening and treatment of cardiovascular disease in patients with SMI [Reference Henderson, Noblett, Parke, Clement, Caffrey and Gale-Grant15, Reference Solmi, Fiedorowicz, Poddighe, Delogu, Miola and Høye16]. Apart from increased rates of alcohol and tobacco consumption [Reference Cook, Wayne, Kafali, Liu, Shu and Flores17], unhealthy nutrition [Reference Teasdale, Ward, Samaras, Firth, Stubbs and Tripodi18], sleep difficulties [Reference Winkelman and Lecea19], physical inactivity, and sedentary behaviour [Reference Vancampfort, Firth, Schuch, Rosenbaum, Mugisha and Hallgren20, Reference Bort-Roig, Briones-Buixassa, Felez-Nobrega, Guàrdia-Sancho, Sitjà-Rabert and Puig-Ribera21] are more prevalent in patients with SMI, and although modifiable, they contribute to the markedly higher prevalence of cardiovascular disease in this clinical population. For example, the diet of adults with SMI is on average less healthy and more obesogenic than the diet of the general population [Reference Firth, Stubbs, Teasdale, Ward, Veronese and Shivappa22] and adults with SMI spend on average about 8 h per day being sedentary during waking hours, which is considerably more than observed in the general population [Reference Vancampfort, Firth, Schuch, Rosenbaum, Mugisha and Hallgren20]. Notably, physical inactivity [Reference Choi, Chen, Stein, Klimentidis, Wang and Koenen23], insomnia [Reference Wang, Cheng and Xu24, Reference Bao, Han, Ma, Wang, Shi and Wang25], and obesity [Reference Tyrrell, Mulugeta, Wood, Zhou, Beaumont and Tuke26–Reference Manu, Khan, Radhakrishnan, Russ, Kane and Correll28] are risk factors contributing not only to chronic cardiovascular diseases, but also to the development or aggravation of mental illnesses. While schizophrenia is typically associated with the greatest degree of sedentary behaviour, unhealthy diet and cardiovascular risk [Reference Vancampfort, Firth, Schuch, Rosenbaum, Mugisha and Hallgren20, Reference Firth, Stubbs, Teasdale, Ward, Veronese and Shivappa22], there is compelling evidence that also people with major depression and bipolar disorder have an increased risk of obesity, diabetes and metabolic syndrome that is at least twice as high as in the general population [Reference Vancampfort, Correll, Galling, Probst, De Hert and Ward29–Reference Vancampfort, Stubbs, Mitchell, De Hert, Wampers and Ward31]. Consequently, the 2018 World Health Organization (WHO) guidelines on the management of physical health conditions in adults with SMI recommend lifestyle interventions as first-line strategies for the prevention of cardiovascular diseases in this population [32]. Lifestyle interventions usually show high acceptability among patients [Reference Erickson, Mena, Pierre, Blum, Martin and Hellemann33] and are increasingly recognized as fundamental for both physical and mental health.
To date, physical activity, including exercise interventions, is the most widely researched lifestyle behaviour in people with SMI and has the potential to improve psychiatric symptoms, cognition, and levels of functioning across a range of mental health diagnoses [Reference Firth, Solmi, Wootton, Vancampfort, Schuch and Hoare34]. These findings were synthesised in a guidance paper by Stubbs et al. published in 2018 [Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi35]. It concluded that physical activity improves depressive symptoms, quality of life and cardiorespiratory fitness in major depressive disorder, and psychiatric symptoms and cognition in schizophrenia spectrum disorders. The current meta-review aims to update this guidance and extend it, considering the efficacy of physical activity interventions when combined with other lifestyle interventions and other lifestyle interventions alone. In general, in the exercise (i.e., structured physical activity) research field, progress is often hampered by inconsistent and imprecise intervention definitions and descriptions, implemented in different target populations and settings with wide variations in the type, frequency, intensity, and duration of interventions in many studies [Reference Seshadri, Adaji, Orth, Singh, Clark and Frye36, Reference Sabe, Kaiser and Sentissi37]. This problem is even more pronounced in other, less researched lifestyle domains [Reference Aucoin, LaChance, Clouthier and Cooley38, Reference De Rosa, Sampogna, Luciano, Del Vecchio, Pocai and Borriello39], which makes it difficult to draw consistent conclusions about the efficacy and effectiveness of interventions and to formulate specific recommendations about their use.
Accordingly, in this guidance paper, we sought to identify the top-tier evidence on lifestyle interventions for adults with SMI to assist in expanding the range of options available to clinicians. This guidance paper consists of two components. The first component is a meta-review of systematic reviews on the effects of non-pharmacological interventions to improve physical activity, diet, and sleep in adults with SMI. We thereby aimed to summarize both their impact on modifiable risk factors for somatic disorders and physical health and to provide new insights about their benefits on mental health outcomes and quality of life. The second component involves using the synthesis of the data to develop a set of practice recommendations and research priorities using the guidance framework of the European Psychiatric Association (EPA). Our goal is to synthesize the existing literature and to mobilize research knowledge into an actionable implementation plan to guide service delivery and to address key research gaps.
Methods
Guidance development process
This meta-review (PROSPERO registration ID CRD42022307336) followed PRISMA guidelines [Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann and Mulrow40] implemented through a pre-determined, published protocol compiled by a research team with different and complementary research expertise within lifestyle-based interventions for mental health. Our team comprises several professional groups, including physicians, psychologists, physiotherapists, sports and exercise scientists, and public health researchers affiliated with institutions from different European countries. The guidance for clinical practice and research was conducted in accordance with the EPA guidelines framework [Reference Gaebel and Möller41].
Definition of severe mental illness
In the literature, different definitions are used to describe SMI. In the present project, we used the definition of the U. S. National Institute of Mental Health (NIMH) of SMI as “a mental, behavioural, or emotional disorder resulting in serious functional impairment, which substantially interferes with or limits one or more major life activities” [6], and is focused on major depression, bipolar disorder and schizophrenia spectrum disorders.
Definition of lifestyle interventions
Currently, there is no consensus definition of the term “lifestyle intervention” which comprises a wide range of interventions. In the medical context, lifestyle interventions usually comprise physical activity, diet, adequate sleep and stress management [Reference Noordsy42]. However, while physical activity, diet and sleep are considered ‘typical’ health behaviours, stress management techniques are a relatively new area of interest where interventions are even less clearly operationalized [Reference Firth, Ward and Stubbs43]. To ensure the feasibility and applicability of this EPA guidance paper, we focused on a narrower definition of lifestyle interventions and restricted the content to non-pharmacological lifestyle interventions focused on physical activity, diet, or sleep interventions. In this meta-review, we use the term physical activity in a broad sense to include interventions to reduce sedentary behaviour and increase structured physical activity, also referred to as exercise or sports interventions. We excluded interventions relating to alcohol or tobacco use, which are often considered psychiatric co-morbidities and have been covered in previous EPA guidance papers and other guidelines on the management of substance use disorders [Reference Rüther, Bobes, De Hert, Svensson, Mann and Batra44, 45]. We included mono- and multimodal interventions, which could have included educational or cognitive-behavioural components, or both, delivered in any setting and in any format.
Systematic search and study identification
Two authors independently searched the databases PubMed, Cochrane Database of Systematic Reviews and Epistemonikos from inception to 16th September 2022 and the database EMBASE from inception to 7 October 2022 for systematic reviews of primary studies investigating lifestyle interventions in adults with SMI. A subsequent updated search for new literature was conducted on 11 March 2024. Since an EPA guidance paper on physical activity for the population of SMI was published in 2018 [Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi35] using overlapping search criteria, the search in the domain of physical activity focused on literature published after the period already covered (i.e., after 16 January 2018). The search was organized in accordance with the population, interventions, comparisons, outcomes, and setting/study (PICOS) reporting structure. Table 1 provides an overview of the search terms included in the overall search strategy, whereas Supplementary material S1 contains the terms for the detailed search strategies adapted to the syntax of different databases. In addition, the reference lists of the included articles were hand-searched to identify further potentially relevant reviews.
Table 1. Application of the PICOS search strategy for the main search terms included in the overall search strategy
Type of studies eligible for inclusion
The inclusion criteria were
-
• systematic reviews with or without meta-analyses that assessed studies of any design;
-
• listed in the databases PubMed, Cochrane Database for Systematic Reviews, Epistemonikos or EMBASE and published in English;
-
• that included adults with schizophrenia spectrum and other psychotic disorders, major depressive disorder or bipolar disorder (with or without cardiovascular comorbidity);
-
• that assessed non-pharmacological lifestyle interventions related to the treatment of mental disorders, targeting physical activity, weight, diet or sleep, including educational or cognitive-behavioural informed interventions, or both, delivered in any setting and in any format;
-
• that included any control;
-
• that considered outcomes regarding targeted health behaviours (e.g., amount of physical activity or reduction of sedentary behaviour), cardiovascular risk factors, psychopathology measures or assessments of the patients’ quality of life or functional outcomes.
We excluded
-
• conference abstracts and studies that were meta-reviews or guidelines that did not employ a systematic search;
-
• systematic reviews of studies in which <50% of participants in the intervention groups had any of the mental disorders under review;
-
• reviews that included people aged younger than 18 years;
-
• reviews that included people that were suffering from disorders other than those defined as SMI here (e.g. seasonal affective disorders, premenstrual dysphoric disorder, peri- and post-partum depression/psychosis) or patients in palliative care;
-
• reviews that assessed pharmacological interventions, stress management (e.g., meditation or mindfulness), dietary interventions solely focused on the intake of specific preparations (e.g., supplements, vitamins) or interventions involving the passive use of objects such as bright light lamps or blue-blocking glasses;
-
• reviews that focused on mechanistic factors of psychopathology as outcomes (e.g. brain structural changes).
-
• reviews considering other disorders or interventions in addition to the specified ones were excluded if the results of the subsets were not reported separately.
Further details on inclusion and exclusion criteria are presented in the Supplementary material S2.
Selection of articles and data extraction
A pair of authors (IM + AS or JS + SM) screened the identified articles on the title and abstract level independently for possible inclusion. Following the initial screen, the full text of potentially eligible reviews was assessed by two authors independently (AS+JS or SM + SW) for eligibility and disagreements were resolved through discussion among all authors (IM, AS, JS, SM, SW). If necessary, review authors were contacted for additional information. The same procedure was applied to potentially eligible articles identified in the reference lists of retrieved articles. Subsequently, we extracted the following data from eligible full-text articles for synthesis: Type of review, search span, study designs included in the review, number of studies included, sample size of the population of interest, mean age of the population, details of the interventions, control groups, outcome measures, main results and conflicts of interest reported.
Quality assessment included systematic reviews and grading of evidence
To assess the certainty of the evidence, we evaluated the quality of the included systematic reviews and meta-analyses using the SIGN Methodology Checklist for Systematic Reviews and Meta-analyses [46] adapted from the AMSTAR tool [Reference Shea, Grimshaw, Wells, Boers, Andersson and Hamel47], as indicated by the EPA guidelines [Reference Heun and Gaebel48]. Accordingly, the quality of each review was rated as high (++), acceptable (+), low (−) or unacceptable (0) (see Table 2). Discrepancies in the ratings were resolved by discussion (AS, JS, SM, SW, IM).
Table 2. Overall assessment of the methodological quality of reviews according to the SIGN Methodology Checklist for Systematic Reviews and Meta-analyses
Guidance development
To develop recommendations, we considered the evidence identified in our systematic review and the quality of reviews as assessed by the SIGN Methodology Checklist [46]. Recommendations were then graded according to the EPA guidance framework (see Table 3) and reviewed by all the authors.
Table 3. Grading of recommendations
Duplicates removed (n = 584)Papers excluded based on title/abstract(n = 2614)Additional records identified through hand searching and screening of lists of references (n = 16)Duplicates removed (n = 584)Papers excluded based on title/abstract(n = 2614)Additional records identified through hand searching and screening of lists of references (n = 16).
Modified from Gaebel W, Großimlinghaus I, Mucic D, Maercker A, Zielasek J, Kerst A. EPA guidance on eMental health interventions in the treatment of posttraumatic stress disorder (PTSD). Eur Psychiatry. 2017;41(1):140–52.
Results
The initial database searches yielded 3353 records and 16 additional records from hand search. After removing duplicates, 2785 records were reviewed at the title and abstract level, of which 2614 were excluded. Subsequently, we assessed 171 full-text articles, of which 106 were excluded (see Figure 1 below). The subsequent updated search for new literature resulted in a further 24 included full-texts. Overall, 89 systematic reviews and meta-analyses were included in this guidance paper. The details of the systematic search process are shown in Figure 1. A list of the excluded full-text reviews of the primary search is given in Supplementary material S3, while a summary of the included reviews is presented in Supplementary materials S4–S7. The number of reviews included in this meta-review is given in Table 4.
Figure 1. PRISMA flowchart (from Page et al. [Reference Page, McKenzie, Bossuyt, Boutron, Hoffmann and Mulrow40]).
Table 4. Number of reviews included in this meta-review
SMI = severe mental illness.
Multimodal lifestyle interventions
Most of the reviews identified cover different types of lifestyle interventions. As primary studies often used combined interventions whose effects were reported jointly, their findings are reported in a separate category on multimodal lifestyle interventions. Details on the included reviews can be found below and in Supplementary material S4.
Severe mental illnesses
We identified a total of 25 systematic reviews (including two updates from previous reviews) focusing on combined lifestyle interventions for mixed populations with SMI. Most of the reviews considered randomized controlled trials (RCTs), while some included also other study types such as quasi-experimental studies.
The interventions of interest consisted of various combinations of physical activity programs [Reference De Rosa, Sampogna, Luciano, Del Vecchio, Pocai and Borriello39, Reference Brown, Geiszler, Lewis and Arbesman49–Reference Stevens, Smith, Bussey, Innerd, McGeechan and Fishburn61], dietary changes including nutrition education and counselling [Reference Aucoin, LaChance, Clouthier and Cooley38, Reference De Rosa, Sampogna, Luciano, Del Vecchio, Pocai and Borriello39, Reference Bruins, Jörg, Bruggeman, Slooff, Corpeleijn and Pijnenborg50–Reference Happell, Davies and Scott65], psychoeducational, motivational or cognitive behavioural techniques to modify behaviour towards healthier lifestyle choices [Reference Aucoin, LaChance, Clouthier and Cooley38, Reference Brown, Geiszler, Lewis and Arbesman49, Reference Caemmerer, Correll and Maayan51, Reference Olker, Parrott, Swarbrick and Spagnolo53–Reference Speyer, Jakobsen, Westergaard, Nørgaard, Jørgensen and Pisinger56, Reference Mohanty, Gandhi, Krishna Prasad, John, Bhaskarapillai and Malo60–Reference McKeon, Papadopoulos, Firth, Joshi, Teasdale and Newby62, Reference Gierisch, Nieuwsma, Bradford, Wilder, Mann-Wrobel and McBroom66–Reference Tosh, Clifton, Xia and White68], peer or family support [Reference Brown, Geiszler, Lewis and Arbesman49, Reference Stubbs, Williams, Shannon, Gaughran and Craig69], social media interventions aiming at increased physical activity or a healthier eating [Reference Brown, Geiszler, Lewis and Arbesman49, Reference Olker, Parrott, Swarbrick and Spagnolo53, Reference Pape, Adriaanse, Kol, van Straten and van Meijel54, Reference McKeon, Papadopoulos, Firth, Joshi, Teasdale and Newby62], or the use of digital technologies targeting lifestyle behaviour or weight loss, such as mobile phone applications or pedometers [Reference Brown, Geiszler, Lewis and Arbesman49, Reference McKeon, Papadopoulos, Firth, Joshi, Teasdale and Newby62, Reference Oliveira, Guariente, Zazula, Mesas, Oliveira and Reiche64].
In summary, interventions combining behaviour change techniques, dietary modification and physical activity were found to be effective in improving weight or body mass index (BMI) in adults with SMI [Reference Brown, Geiszler, Lewis and Arbesman49–Reference Caemmerer, Correll and Maayan51, Reference Olker, Parrott, Swarbrick and Spagnolo53, Reference Singh, Karmani, Malo, Virupaksha, Muralidhar and Venkatasubramanian55, Reference Speyer, Jakobsen, Westergaard, Nørgaard, Jørgensen and Pisinger56, Reference Meader, Melton, Evans, Wright, Shiers and Ratschen58, Reference Lee, Piernas, Stewart, Michalopoulou, Hajzadeh and Edwards59, Reference Happell, Davies and Scott65–Reference Verhaeghe, De Maeseneer, Maes, Van Heeringen and Annemans67]. However, the clinical relevance of findings was criticised repeatedly in the systematic reviews (usually only a weight loss of more than 5% is considered as clinically relevant to reduce health-related complications) [Reference Olker, Parrott, Swarbrick and Spagnolo53, Reference Speyer, Jakobsen, Westergaard, Nørgaard, Jørgensen and Pisinger56, Reference Verhaeghe, De Maeseneer, Maes, Van Heeringen and Annemans67], and more consistent results were found, especially for interventions with a study duration of more than 6 months [Reference Naslund, Whiteman, McHugo, Aschbrenner, Marsch and Bartels52, Reference Taylor, Stubbs, Hewitt, Ajjan, Alderson and Gilbody57]. Several effective elements of these behavioural interventions were identified, including regular contact, supporting tools, such as intervention handbooks or pedometers, and educational materials [Reference Lee, Piernas, Stewart, Michalopoulou, Hajzadeh and Edwards59]. In addition, multimodal interventions were shown to decrease distinct aspects of the metabolic syndrome, such as fasting glucose levels [Reference Bruins, Jörg, Bruggeman, Slooff, Corpeleijn and Pijnenborg50, Reference Caemmerer, Correll and Maayan51, Reference Taylor, Stubbs, Hewitt, Ajjan, Alderson and Gilbody57] or triglyceride levels [Reference Bruins, Jörg, Bruggeman, Slooff, Corpeleijn and Pijnenborg50, Reference Caemmerer, Correll and Maayan51].
Interventions using digital technologies (such as pedometers or mobile phone communication) have been reported to be easy to use and to reduce sedentary behaviour [Reference Oliveira, Guariente, Zazula, Mesas, Oliveira and Reiche64]. Social media interventions were also considered highly acceptable, but only resulted in non-significant improvements in cardiorespiratory fitness [Reference McKeon, Papadopoulos, Firth, Joshi, Teasdale and Newby62]. The evidence on the effects of peer or family support remains inconclusive [Reference Stubbs, Williams, Shannon, Gaughran and Craig69]. Most reviews focused on improvements in physical health, and mental health outcomes were less consistently reported. Reviews that examined effects on quality of life showed heterogeneous results [Reference Aucoin, LaChance, Clouthier and Cooley38, Reference Pape, Adriaanse, Kol, van Straten and van Meijel54, Reference Speyer, Jakobsen, Westergaard, Nørgaard, Jørgensen and Pisinger56, Reference Meader, Melton, Evans, Wright, Shiers and Ratschen58, Reference Stevens, Smith, Bussey, Innerd, McGeechan and Fishburn61, Reference Verhaeghe, De Maeseneer, Maes, Van Heeringen and Annemans67, Reference Tosh, Clifton, Xia and White68], whereas effects on depressive symptoms [Reference Bruins, Jörg, Bruggeman, Slooff, Corpeleijn and Pijnenborg50, Reference Pape, Adriaanse, Kol, van Straten and van Meijel54] or anxiety [Reference Pape, Adriaanse, Kol, van Straten and van Meijel54] were more consistently detected. In general, delivering interventions both in a group setting as well as in individual sessions has been recommended [Reference Aucoin, LaChance, Clouthier and Cooley38, Reference Speyer, Jakobsen, Westergaard, Nørgaard, Jørgensen and Pisinger56]. The quality of the included reviews was variable, ranging from unacceptable to high [Reference Pape, Adriaanse, Kol, van Straten and van Meijel54, Reference Speyer, Jakobsen, Westergaard, Nørgaard, Jørgensen and Pisinger56, Reference Taylor, Stubbs, Hewitt, Ajjan, Alderson and Gilbody57, Reference Lee, Piernas, Stewart, Michalopoulou, Hajzadeh and Edwards59, Reference Stevens, Smith, Bussey, Innerd, McGeechan and Fishburn61, Reference McKeon, Papadopoulos, Firth, Joshi, Teasdale and Newby62, Reference Tosh, Clifton, Xia and White68].
Recommendation (Grade A)
Lifestyle interventions that combine behaviour change techniques, dietary modifications, and physical activity should be offered to individuals with SMI for reducing weight and other metabolic syndrome components, as well as depressive and anxiety symptoms.
Major depressive disorder
We did not find any systematic reviews that separately investigated the effects of mixed lifestyle interventions in adults with major depressive disorder.
Bipolar disorder
Three systematic reviews reporting data from 1898 participants considered individuals with bipolar disorder [Reference Bauer, Gálvez, Hamilton, Balanzá-Martínez, Zunta-Soares and Soares70–Reference Simjanoski, Patel, Boni, Balanzá-Martínez, Frey and Minuzzi72]. The reviews included RCTs [Reference Simjanoski, Patel, Boni, Balanzá-Martínez, Frey and Minuzzi72] as well as non-randomized controlled studies [Reference Bauer, Gálvez, Hamilton, Balanzá-Martínez, Zunta-Soares and Soares70, Reference Gliddon, Barnes, Murray and Michalak71]. Two reviews focused on multimodal lifestyle intervention programs covering domains such as dietary modification, physical activity, sleep, and motivational or self-management aspects, but identified few primary studies that met the inclusion criteria [Reference Bauer, Gálvez, Hamilton, Balanzá-Martínez, Zunta-Soares and Soares70, Reference Simjanoski, Patel, Boni, Balanzá-Martínez, Frey and Minuzzi72]. In the systematic review by Bauer et al., weight or BMI was the primary outcome in most of the trials. Overall, results pointed towards beneficial effects on weight, blood pressure, lipid profile, physical activity and mood, but the quality of the underlying studies was criticized due to small sample sizes, variability in assessments and an insufficient number of RCTs [Reference Bauer, Gálvez, Hamilton, Balanzá-Martínez, Zunta-Soares and Soares70]. The quality of the review was rated as acceptable according to SIGN. The systematic review by Simjanoski et al. concluded that combined interventions covering diet, physical activity and sleep reduced depressive symptoms and improved overall functioning [Reference Simjanoski, Patel, Boni, Balanzá-Martínez, Frey and Minuzzi72]. The quality of this review was rated as high.
Another review examined the effects of e-Health (health services and information via the internet) and m-Health (use of mobile devices) interventions to improve self-management, which included maintaining a healthy lifestyle as well as topics such as mood or medication monitoring, or self-education about the illness [Reference Gliddon, Barnes, Murray and Michalak71]. However, only few primary studies incorporated lifestyle aspects. Review quality was rated as low.
Recommendation (Grade B)
Multimodal lifestyle interventions can be used to improve overall functioning and reduce depressive symptoms as well as cardiovascular risk factors in adults with bipolar disorder.
Schizophrenia spectrum disorders
A total of 15 systematic reviews summarized the effects of multimodal lifestyle interventions in adults with schizophrenia spectrum disorders. Most reviews included RCTs [Reference Alvarez-Jimenez, Hetrick, Gonzalez-Blanch, Gleeson and McGorry73–Reference Coles, Maksyutynska, Knezevic, Agarwal, Strudwick and Dunbar79], while some also included studies with non-randomized designs, such as quasi-experimental trials, case studies, cohort or retrospective studies [Reference Faulkner, Soundy and Lloyd80–Reference Cimo, Stergiopoulos, Cheng, Bonato and Dewa82].
The interventions of interest mostly involved weight management programs that incorporated nutritional education and counselling or caloric restriction [Reference Alvarez-Jimenez, Hetrick, Gonzalez-Blanch, Gleeson and McGorry73–Reference Fernández-Abascal, Suárez-Pinilla, Cobo-Corrales, Crespo-Facorro and Suárez-Pinilla75, Reference Mucheru, Hanlon, McEvoy, Thakkinstian and MacDonald-Wicks77, Reference Faulkner, Soundy and Lloyd80–Reference Loh, Meyer and Leckband87], physical activity promotion or exercise programs [Reference Alvarez-Jimenez, Hetrick, Gonzalez-Blanch, Gleeson and McGorry73–Reference Zavala, Todowede, Mazumdar, Aslam, Choudhury and Jarde78, Reference Faulkner, Soundy and Lloyd80–Reference Cimo, Stergiopoulos, Cheng, Bonato and Dewa82, Reference Tumiel, Wichniak, Jarema and Lew-Starowicz84, Reference Whitney, Procyshyn, Fredrikson and Barr86], and cognitive behavioural or motivational elements (such as social reinforcement, token economy, goal setting, progress monitoring) [Reference Alvarez-Jimenez, Hetrick, Gonzalez-Blanch, Gleeson and McGorry73–Reference Mucheru, Hanlon, McEvoy, Thakkinstian and MacDonald-Wicks77, Reference Hjorth, Davidsen, Kilian and Skrubbeltrang81, Reference Tumiel, Wichniak, Jarema and Lew-Starowicz84, Reference Loh, Meyer and Leckband87].
All the reviews assessed the effects of lifestyle interventions on body weight or BMI as the main outcome, and all found small to moderate improvements. Other cardiometabolic risk factors, such as waist circumference [Reference Fernández-Abascal, Suárez-Pinilla, Cobo-Corrales, Crespo-Facorro and Suárez-Pinilla75, Reference Mazoruk, Meyrick, Taousi and Huxley76], blood glucose levels [Reference Cimo, Stergiopoulos, Cheng, Bonato and Dewa82] or cardiorespiratory fitness [Reference Tumiel, Wichniak, Jarema and Lew-Starowicz84] were also improved. Peer support was found to increase the amount of physical activity in daily life and to improve cardiorespiratory fitness [Reference Coles, Maksyutynska, Knezevic, Agarwal, Strudwick and Dunbar79].
Mental health outcomes were less commonly reported. Effects on quality of life were inconsistent [Reference Alvarez-Jimenez, Hetrick, Gonzalez-Blanch, Gleeson and McGorry73, Reference Fernández-Abascal, Suárez-Pinilla, Cobo-Corrales, Crespo-Facorro and Suárez-Pinilla75, Reference Zavala, Todowede, Mazumdar, Aslam, Choudhury and Jarde78], while beneficial effects on positive, negative or general symptoms and cognition were shown particularly in programs incorporating physical activity elements [Reference Fernández-Abascal, Suárez-Pinilla, Cobo-Corrales, Crespo-Facorro and Suárez-Pinilla75, Reference Mazoruk, Meyrick, Taousi and Huxley76]. The quality of the included reviews ranged from unacceptable to high quality [Reference Alvarez-Jimenez, Hetrick, Gonzalez-Blanch, Gleeson and McGorry73–Reference Zavala, Todowede, Mazumdar, Aslam, Choudhury and Jarde78, Reference Cimo, Stergiopoulos, Cheng, Bonato and Dewa82].
These findings collectively suggest that programs combining both physical activity, nutritional counselling and motivational or behavioural change techniques may be the most successful.
Recommendation (Grade A)
To improve healthy weight management in individuals with schizophrenia spectrum disorders, multicomponent programs that include physical activity, nutritional counselling and motivational and/or cognitive behavioural techniques should be offered.
Physical activity
In total, we identified 36 systematic reviews dealing with the effects of physical activity as the sole intervention. Details on the included reviews are given below and in Supplementary material S5.
Severe mental illnesses
In our systematic search to complement the recent EPA guidance paper on physical activity in SMI [Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi35], we found seven new systematic reviews that included mixed SMI populations [Reference Ashdown-Franks, Williams, Vancampfort, Firth, Schuch and Hubbard88–Reference Yu, Wong, Lei, Nie, Shi and Zou94], five of which conducted a meta-analysis.
Three reviews included only RCTs [Reference Brinsley, Schuch, Lederman, Girard, Smout and Immink90, Reference Tavares, Rossell, Schuch, Herring, Menezes de Sousa and Galvão-Coelho92, Reference Yu, Wong, Lei, Nie, Shi and Zou94], while the remainder also considered non-randomized controlled trials as well as pre-post intervention studies [Reference Ashdown-Franks, Williams, Vancampfort, Firth, Schuch and Hubbard88, Reference Korman, Armour, Chapman, Rosenbaum, Kisely and Suetani89, Reference Martland, Korman, Firth and Stubbs91, Reference Sawyer, McKeon, Hassan, Onyweaka, Martinez Agulleiro and Guinart93]. Four reviews studied a variety of physical activity interventions [Reference Ashdown-Franks, Williams, Vancampfort, Firth, Schuch and Hubbard88, Reference Martland, Korman, Firth and Stubbs91, Reference Tavares, Rossell, Schuch, Herring, Menezes de Sousa and Galvão-Coelho92, Reference Yu, Wong, Lei, Nie, Shi and Zou94]. Others addressed specific types of exercise, including yoga [Reference Brinsley, Schuch, Lederman, Girard, Smout and Immink90] and high-intensity interval training (HIIT) [Reference Korman, Armour, Chapman, Rosenbaum, Kisely and Suetani89] or digital interventions [Reference Sawyer, McKeon, Hassan, Onyweaka, Martinez Agulleiro and Guinart93]. One review focused only on inpatients [Reference Martland, Korman, Firth and Stubbs91]. To date, it is unclear whether physical activity interventions are effective in improving physical health parameters [Reference Martland, Korman, Firth and Stubbs91] or in increasing general physical activity levels in the daily lives of individuals with an SMI [Reference Ashdown-Franks, Williams, Vancampfort, Firth, Schuch and Hubbard88]. Similarly, digital health behaviour interventions showed mixed results [Reference Sawyer, McKeon, Hassan, Onyweaka, Martinez Agulleiro and Guinart93]. With respect to depressive symptoms, yoga appears to be more effective than various inactive control conditions such as educative sessions, waitlists or TAU [Reference Brinsley, Schuch, Lederman, Girard, Smout and Immink90]. High-intensity interval training may also be a highly acceptable and safe form of exercise with potential antidepressant effects [Reference Korman, Armour, Chapman, Rosenbaum, Kisely and Suetani89]. Another systematic review concluded that especially multimodal exercise interventions are beneficial in relieving depressive symptoms [Reference Yu, Wong, Lei, Nie, Shi and Zou94].
Moreover, physical activity has been shown to have positive effects on several cognitive domains, such as reasoning and problem solving [Reference Tavares, Rossell, Schuch, Herring, Menezes de Sousa and Galvão-Coelho92]. Physical activity interventions that aim to achieve at least a moderately intense level of physical exertion [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95] and a minimum of 150 min per week can generally be recommended to improve the mental health condition in individuals with SMI [5].
The quality of individual reviews according to SIGN ranged from unacceptable to high quality [Reference Martland, Korman, Firth and Stubbs91, Reference Tavares, Rossell, Schuch, Herring, Menezes de Sousa and Galvão-Coelho92, Reference Yu, Wong, Lei, Nie, Shi and Zou94]. The recommendation from the previous EPA guidance paper on physical activity [Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi35] can be updated as follows:
Recommendation (Grade A)
Physical activity should be offered to individuals with SMI for improving depressive symptoms and cognitive functioning.
Major depressive disorder
Nine recent reviews, each supplemented with a meta-analysis, were included focusing on major depressive disorder [Reference Brupbacher, Gerger, Zander-Schellenberg, Straus, Porschke and Gerber96–Reference Ren, Hillman, Wang, Li, Zhou and Liang104]. The primary studies included in the reviews were almost exclusively RCTs of physical exercise in general [Reference Chen and Shan97], aerobic exercise [Reference Ren, Hillman, Wang, Li, Zhou and Liang104], or aerobic and strength training with [Reference Brupbacher, Gerger, Zander-Schellenberg, Straus, Porschke and Gerber96] or without [Reference Heissel, Heinen, Brokmeier, Skarabis, Kangas and Vancampfort101, Reference Pérez Bedoya, Puerta-López, López Galvis, Rojas Jaimes and Moreira103] meditative movements. One review investigated yoga interventions only [Reference Wu, Yan and Yang102], one reviewed qigong-based therapy [Reference Guo, Kong and Zhang98], while another distinguished between exercise, yoga, and tai chi [Reference Seshadri, Adaji, Orth, Singh, Clark and Frye99]. Despite some null findings [Reference Pérez Bedoya, Puerta-López, López Galvis, Rojas Jaimes and Moreira103], aerobic and resistance exercise appears to be superior in improving depressive symptoms [Reference Seshadri, Adaji, Orth, Singh, Clark and Frye99–Reference Heissel, Heinen, Brokmeier, Skarabis, Kangas and Vancampfort101] when compared to various control conditions [Reference Chen and Shan97]. Moreover, aerobic exercise interventions showed beneficial effects on overall cognitive functioning and the subdomains of memory and executive function [Reference Ren, Hillman, Wang, Li, Zhou and Liang104]. In addition, different types of physical activity may improve sleep quality, with mind–body exercise or vigorous strength training [Reference Brupbacher, Gerger, Zander-Schellenberg, Straus, Porschke and Gerber96] showing the most promising results. Add-on yoga may also improve depressive symptoms [Reference Wu, Yan and Yang102], although the quality of evidence was low to moderate [Reference Wu, Yan and Yang102], and the effect was reduced when only studies with a low risk of bias were considered [Reference Seshadri, Adaji, Orth, Singh, Clark and Frye99]. To date, evidence for an antidepressant effect of other mind–body exercise forms, such as tai chi [Reference Seshadri, Adaji, Orth, Singh, Clark and Frye99] and qigong [Reference Guo, Kong and Zhang98], is insufficient. Further, no conclusions can be drawn about the physical health benefits and effects on the quality of life [Reference Pérez Bedoya, Puerta-López, López Galvis, Rojas Jaimes and Moreira103] of physical activity in individuals with major depressive disorder. Except for two reviews, we rated the quality of reviews as acceptable to high [Reference Brupbacher, Gerger, Zander-Schellenberg, Straus, Porschke and Gerber96, Reference Seshadri, Adaji, Orth, Singh, Clark and Frye99, Reference Heissel, Heinen, Brokmeier, Skarabis, Kangas and Vancampfort101–Reference Ren, Hillman, Wang, Li, Zhou and Liang104]. Accordingly, the updated recommendation is as follows:
Recommendation (Grade A)
Physical activity should be offered as a treatment to improve depressive symptoms, cognitive functioning, and sleep quality in adults with major depressive disorder. Among meditative forms of physical activity, the strength of evidence is greatest for yoga.
Bipolar disorder
As with the previous guidance paper on physical activity [Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi35], no systematic reviews could be identified focusing on the effects of physical activity in adults with bipolar disorder. To fill this research gap, future work should aim to synthesise the existing findings in this population.
Schizophrenia spectrum disorders
Twenty recent systematic reviews reported on the effects of physical activity in individuals with schizophrenia spectrum disorders, 16 of which included a meta-analysis [Reference Sabe, Kaiser and Sentissi37, Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105–Reference Guo, Liu, Liao, Qin and Yue118]. Twelve reviews analysed only RCTs [Reference Sabe, Kaiser and Sentissi37, Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Vogel, van der Gaag, Slofstra, Knegtering, Bruins and Castelein107, Reference Yip, Karatzias and Chien108, Reference Ziebart, Bobos, MacDermid, Furtado, Sobczak and Doering110, Reference Sabe, Sentissi and Kaiser112, Reference Yin, Sun, Zhu, Alifujiang, Wang and An114, Reference Kim, Lee and Kang115, Reference Su, Pan, Li and Zhang117–Reference Chang, Goh, Latthirun and Yang120], and the remainder also included non-randomized controlled trials [Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105, Reference Swora, Boberska, Kulis, Knoll, Keller and Luszczynska106, Reference Xu, Cai, Fang, Zheng, Shan and Yang109, Reference Bredin, Kaufman, Chow, Lang, Wu and Kim111, Reference Gallardo-Gómez, Noetel, Álvarez-Barbosa, Alfonso-Rosa, Ramos-Munell and Del Pozo Cruz113, Reference Oliva, Monteiro-Junior, Oliva and Powers116, Reference Shannon, McGuire, Brown and O’Donoghue121, Reference Cempa, Jurys, Kluczyński and Andreew122].
Interventions were defined as any form of physical activity [Reference Sabe, Kaiser and Sentissi37, Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Swora, Boberska, Kulis, Knoll, Keller and Luszczynska106, Reference Ziebart, Bobos, MacDermid, Furtado, Sobczak and Doering110, Reference Gallardo-Gómez, Noetel, Álvarez-Barbosa, Alfonso-Rosa, Ramos-Munell and Del Pozo Cruz113, Reference Kim, Lee and Kang115, Reference Oliva, Monteiro-Junior, Oliva and Powers116, Reference Guo, Liu, Liao, Qin and Yue118, Reference Chang, Goh, Latthirun and Yang120, Reference Cempa, Jurys, Kluczyński and Andreew122], aerobic or resistance exercise [Reference Sabe, Kaiser and Sentissi37, Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105, Reference Vogel, van der Gaag, Slofstra, Knegtering, Bruins and Castelein107, Reference Xu, Cai, Fang, Zheng, Shan and Yang109, Reference Bredin, Kaufman, Chow, Lang, Wu and Kim111], and mind–body exercises such as yoga and tai chi [Reference Vogel, van der Gaag, Slofstra, Knegtering, Bruins and Castelein107, Reference Yip, Karatzias and Chien108, Reference Sabe, Sentissi and Kaiser112, Reference Yin, Sun, Zhu, Alifujiang, Wang and An114, Reference Su, Pan, Li and Zhang117]. Two reviews focused on the mode of intervention delivery, including group-based [Reference Shannon, McGuire, Brown and O’Donoghue121] and videogame-based exercise [Reference Roberts, Lloyd, Välimäki, Ho, Freemantle and Bekefi119].
Physical activity interventions in general were shown to be effective in alleviating positive [Reference Kim, Lee and Kang115, Reference Oliva, Monteiro-Junior, Oliva and Powers116], general [Reference Oliva, Monteiro-Junior, Oliva and Powers116], and especially negative symptoms of schizophrenia spectrum disorders [Reference Sabe, Kaiser and Sentissi37, Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Vogel, van der Gaag, Slofstra, Knegtering, Bruins and Castelein107, Reference Bredin, Kaufman, Chow, Lang, Wu and Kim111, Reference Gallardo-Gómez, Noetel, Álvarez-Barbosa, Alfonso-Rosa, Ramos-Munell and Del Pozo Cruz113, Reference Oliva, Monteiro-Junior, Oliva and Powers116, Reference Guo, Liu, Liao, Qin and Yue118, Reference Chang, Goh, Latthirun and Yang120] as well as the quality of life [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95]. Moreover, improvements in cognitive functioning and functioning in daily life have been found [Reference Chang, Goh, Latthirun and Yang120]. Effects on depressive symptoms were mixed [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Kim, Lee and Kang115]. While the effects of resistance exercise or exercise combinations, including resistance exercise on PANSS scores were limited by the small number of studies [Reference Vogel, van der Gaag, Slofstra, Knegtering, Bruins and Castelein107, Reference Bredin, Kaufman, Chow, Lang, Wu and Kim111], there is particularly strong evidence for aerobic exercise [Reference Sabe, Kaiser and Sentissi37, Reference Vogel, van der Gaag, Slofstra, Knegtering, Bruins and Castelein107, Reference Ziebart, Bobos, MacDermid, Furtado, Sobczak and Doering110, Reference Bredin, Kaufman, Chow, Lang, Wu and Kim111]. Further, aerobic exercise can reduce PANSS positive and general symptoms [Reference Ziebart, Bobos, MacDermid, Furtado, Sobczak and Doering110, Reference Bredin, Kaufman, Chow, Lang, Wu and Kim111] and improve cognitive functioning [Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105, Reference Xu, Cai, Fang, Zheng, Shan and Yang109]. Enhancements in global cognition and several cognitive subdomains appear to be particularly supported for group-based [Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105, Reference Shannon, McGuire, Brown and O’Donoghue121] and supervised [Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105, Reference Xu, Cai, Fang, Zheng, Shan and Yang109] aerobic interventions. Overall, benefits were found regardless of the type of exercise performed [Reference Gallardo-Gómez, Noetel, Álvarez-Barbosa, Alfonso-Rosa, Ramos-Munell and Del Pozo Cruz113], including interventions with mind–body exercise components [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95] and multicomponent interventions [Reference Kim, Lee and Kang115].
Mind–body exercise, such as yoga and tai chi, may also be more effective in reducing negative symptoms than various control conditions [Reference Vogel, van der Gaag, Slofstra, Knegtering, Bruins and Castelein107, Reference Sabe, Sentissi and Kaiser112, Reference Yin, Sun, Zhu, Alifujiang, Wang and An114, Reference Su, Pan, Li and Zhang117]. However, there is currently a lack of specific evidence for tai chi [Reference Sabe, Sentissi and Kaiser112]. With respect to yoga interventions, three reviews reported beneficial effects of yoga compared to treatment as usual on positive [Reference Yip, Karatzias and Chien108, Reference Su, Pan, Li and Zhang117] or negative symptoms [Reference Yin, Sun, Zhu, Alifujiang, Wang and An114, Reference Su, Pan, Li and Zhang117]. However, another review found no significant differences [Reference Ziebart, Bobos, MacDermid, Furtado, Sobczak and Doering110], while a fourth reported that positive effects of yoga on these symptoms were no longer significant after excluding studies with a high risk of bias or outliers [Reference Sabe, Sentissi and Kaiser112]. In general, reported effects of interventions including mind–body exercises were more heterogeneous than those without, potentially indicating higher instructor-dependence [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95].
No consistent effects of physical activity interventions on body weight could be detected [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Guo, Liu, Liao, Qin and Yue118].
There is still insufficient research on the optimal delivery of physical activity interventions. While there is only very limited evidence for videogame-based training [Reference Roberts, Lloyd, Välimäki, Ho, Freemantle and Bekefi119], we found the most support for professionally supervised group exercise [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105, Reference Shannon, McGuire, Brown and O’Donoghue121]. In addition, higher levels of physical activity generally appear to be associated with increasing benefits for positive, negative and general symptoms [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95, Reference Swora, Boberska, Kulis, Knoll, Keller and Luszczynska106], but different recommendations have been made in terms of the minimum duration and intensity required [Reference Sabe, Kaiser and Sentissi37, Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105]. While some reviews suggested shorter durations [Reference Shimada, Ito, Makabe, Yamanushi, Takenaka and Kawano105, Reference Guo, Liu, Liao, Qin and Yue118], as general agreement at least moderate-intensity exercise [Reference Sabe, Kaiser and Sentissi37, Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95] should be achieved in sessions lasting more than 120 min per week [Reference Sabe, Kaiser and Sentissi37, Reference Guo, Liu, Liao, Qin and Yue118] (150 min per week according to the WHO [5]) for more than 6 months [Reference Sabe, Kaiser and Sentissi37, Reference Kim, Lee and Kang115]. In general, more frequent sessions seem to be more effective in improving psychosocial functioning [Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95].
Albeit the quality of the primary research was repeatedly criticized, the reviews were almost exclusively of high quality. Therefore, the recommendation from the previous EPA guidance paper [Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi35] can be updated as follows:
Recommendation (Grade A)
Physical activity should be offered to individuals with schizophrenia spectrum disorders as an adjunctive treatment to improve positive, negative and general psychopathology symptoms, cognition, and quality of life.
Diet
Eight different systematic reviews were found investigating dietary interventions. Details on the included reviews are given below and in Supplementary material S6.
Severe mental illnesses
Three of the eight systematic reviews focused on dietary interventions for individuals with SMI [Reference Aranburu, Matias, Simón, Larretxi, Martínez and Bustamante123–Reference Rocks, Teasdale, Fehily, Howland and Kelly125]. Due to the combined reporting of results from clinical populations with depression and anxiety, we could consider only one study (n = 16) conducted in individuals with schizophrenia from the systematic review by Aranburu et al. [Reference Aranburu, Matias, Simón, Larretxi, Martínez and Bustamante123], which examined the effects of a gluten-free versus gluten-containing diet. The dietary interventions reported by Teasdale et al. consisted of individualized counselling, group education, or a combination of these [Reference Teasdale, Ward, Rosenbaum, Samaras and Stubbs124]. Significant reductions in weight, BMI, and waist circumference were reported compared to treatment as usual, leading the authors to conclude that dietary interventions may be beneficial for weight control in SMI, particularly when delivered by dietitians [Reference Teasdale, Ward, Rosenbaum, Samaras and Stubbs124]. The systematic review by Rocks et al. also emphasises the importance of interventions delivered by dietitians and in individual sessions, but found only limited evidence for the effectiveness of dietary interventions in improving metabolic syndrome risk factors in people with SMI in general [Reference Rocks, Teasdale, Fehily, Howland and Kelly125]. The quality of this review was rated as high [Reference Rocks, Teasdale, Fehily, Howland and Kelly125], while the other two reviews were rated as low or unacceptable. Due to the limited amount and quality of the available evidence, no firm conclusions can currently be drawn from these reviews for adults with SMI.
Major depressive disorder
We identified three systematic reviews that investigated the impact of dietary interventions on depressive symptoms. The interventions included a 6-month Mediterranean diet [Reference Altun, Brown, Szoeke and Goodwill126], different forms of caloric or carbohydrate restriction [Reference Swainson, Reeson, Malik, Stefanuk, Cummins and Sivapalan127], or Intermittent (Ramadan) Fasting [Reference Murta, Seixas, Harada, Damiano and Zanetti128]. Intermittent Fasting had a beneficial effect on anthropometric outcomes, but the review quality was rated as low. Both the Mediterranean diet or caloric restriction interventions appeared to be associated with fewer depressive episodes [Reference Altun, Brown, Szoeke and Goodwill126] or mood improvements [Reference Swainson, Reeson, Malik, Stefanuk, Cummins and Sivapalan127] when compared with the control conditions. However, only one review was of acceptable quality [Reference Swainson, Reeson, Malik, Stefanuk, Cummins and Sivapalan127], and there was considerable heterogeneity between studies, which limits the generalisability of the results. Further interventional research is therefore needed to formulate more robust dietary recommendations for patients with major depressive disorder.
Recommendation (Grade D)
Dietary interventions may be considered to improve depressive symptoms in adults with major depressive disorder.
Bipolar disorder
One systematic review aimed to examine the effectiveness of obesity interventions in adults with bipolar disorder, but found no studies that met the inclusion criteria, highlighting the need for research in this area [Reference Tully, Smyth, Conway, Geddes, Devane and Kelly129].
Schizophrenia spectrum disorders
Only one systematic review was identified that aimed to investigate dietary interventions as a single lifestyle modification in adults with schizophrenia [Reference Pearsall, Praveen, Pelosi and Geddes130]. However, the authors did not find any studies that met the inclusion criteria [Reference Pearsall, Praveen, Pelosi and Geddes130]. Therefore, currently, no recommendation can be made regarding dietary interventions alone in adults with schizophrenia spectrum disorders. Future experimental studies are needed to identify the most beneficial dietary patterns and food components in this population.
Sleep interventions
We found two systematic reviews that investigated sleep interventions. Details on these reviews are provided below and in the Supplementary material S7.
Severe mental illnesses
There were no systematic reviews that summarised the results of sleep interventions in mixed SMI populations.
Major depressive disorder
We did not find any systematic reviews of sleep interventions for adults with major depressive disorder.
Bipolar disorder
One review investigated non-pharmacological interventions targeting sleep and circadian rhythms in adults with bipolar disorder and included 10 primary RCTs [Reference Bisdounis, Saunders, Farley, Lee, McGowan and Espie131]. Interventions were heterogeneous, ranging from sleep deprivation (n = 1), interpersonal and social rhythm therapy to improve sleep behaviour (n = 4), cognitive behavioural therapy-informed intervention for insomnia (n = 1) and combination treatments (n = 4). In sum, interpersonal and social rhythm therapy to improve sleep behaviour yielded conflicting results on mood symptoms, while total sleep deprivation and combination therapies showed no effect. The only available study on cognitive behavioural therapy-informed intervention for insomnia found significant reductions in depression scores, improvements in insomnia symptoms and improved sleep quality. Of note, only a minority of primary studies included sleep or circadian rhythm as an outcome, focusing instead on affective symptoms [Reference Bisdounis, Saunders, Farley, Lee, McGowan and Espie131]. Although the review was rated as high quality according to SIGN criteria, it provides only limited evidence for the effectiveness of the interventions in question due to the small number of studies per intervention, as well as the heterogeneity of intervention protocols, outcomes and patient characteristics. Thus, the current evidence does not allow firm conclusions to be drawn about the effects of sleep interventions in adults with bipolar disorder.
Schizophrenia spectrum disorders
We identified one systematic review [Reference Waite, Sheaves, Isham, Reeve and Freeman132] investigating the effectiveness of psychological interventions for sleep disturbances in adults with schizophrenia, of which a subset of four primary studies with a total of 263 participants was relevant to the present work. The review included RCTs and non-randomized trials with treatment as usual control groups and focused on cognitive behavioural therapy-informed interventions targeting insomnia or, in one case, nightmares. Treatment duration ranged from two to 12 weeks. The interventions led to sleep improvements, with large effect sizes post-treatment and at follow-up. The results also indicated modest improvements in positive symptoms, but the trials were not adequately powered to detect this effect in patient populations. Moreover, the quality of this review was rated as unacceptable according to SIGN criteria because its literature search was limited to one electronic database; therefore, no clear recommendations for the use of sleep interventions in adults with schizophrenia spectrum disorders can be made at this time based on the current evidence. More clinical trials, which should include outcomes of sleep and circadian rhythm in addition to psychopathological outcomes, are needed to determine the most beneficial intervention protocols.
Recommendation (Grade D)
Cognitive behavioural therapy-informed interventions targeting insomnia may be considered to improve sleep quality and mental health outcomes in adults with SMI.
A summary of all recommendation grades is given in Table 5.
Table 5. Summary of recommendation grades
SMI = severe mental illness. Definitions of the recommendation grades are given in Table 3. A minus indicates a lack of systematic reviews (or their insufficient quality) to formulate a recommendation.
Discussion
In this EPA Guidance paper, we conducted a meta-review to synthesise the evidence on lifestyle interventions for adults with SMI and formulate recommendations for both clinical practice and research in this area.
Many of the identified reviews included the mixed diagnosis group of SMI. In summary, we found that lifestyle interventions that combine behavioural change techniques, dietary modification and physical activity show the greatest promise in reducing weight and improving additional cardiovascular health parameters. Furthermore, in line with the previous research [Reference Firth, Solmi, Wootton, Vancampfort, Schuch and Hoare34], we conclude that add-on physical activity can be used to improve also depressive symptoms in adults with SMI. Even if there is still insufficient evidence for the optimal delivery of exercise interventions, group-based sessions supervised by qualified professionals [Reference Vancampfort, Rosenbaum, Schuch, Ward, Probst and Stubbs133, Reference Firth, Siddiqi, Koyanagi, Siskind, Rosenbaum and Galletly134] of at least moderate intensity [Reference Sabe, Kaiser and Sentissi37, Reference Rißmayer, Kambeitz, Javelle and Lichtenstein95] for a total of more than 150 min per week [5] seem to be advantageous.
In adults with major depressive disorder, physical activity was shown to ameliorate depressive symptoms and improve cognitive functioning, sleep quality and physical fitness. A recent umbrella review also found significant effects of exercise interventions on physical health outcomes in this population [Reference Croatto, Vancampfort, Miola, Olivola, Fiedorowicz and Firth135], and associations between dietary interventions and reduced depressive symptoms have also been reported [Reference Altun, Brown, Szoeke and Goodwill126, Reference Swainson, Reeson, Malik, Stefanuk, Cummins and Sivapalan127]. In general, as noted before [Reference Stubbs, Vancampfort, Hallgren, Firth, Veronese and Solmi35], there is still a lack of reviews on the benefits of lifestyle interventions for individuals with bipolar disorder. Nevertheless, multimodal lifestyle interventions can be a promising approach, offering potential benefits for improved cardiovascular risk factors, overall functioning and reduced depressive symptoms in people with bipolar disorder.
Most meta-research focused on lifestyle interventions in adults with schizophrenia spectrum disorders. Effective weight loss programs are multicomponent and incorporate physical activity, nutritional counselling, and motivational and/or cognitive behavioural techniques. In terms of mental health outcomes, physical activity is a useful adjunct treatment to improve schizophrenia-related positive, negative and general psychopathology, cognition and quality of life, which is in line with the previous literature [Reference Firth, Solmi, Wootton, Vancampfort, Schuch and Hoare34]. Moreover, cognitive behavioural therapy-informed interventions can be considered to ameliorate sleep quality in adults with schizophrenia.
Our recommendations underscore the importance of lifestyle interventions in improving the physical and mental health and overall well-being of individuals with SMI. The strength of evidence varies across interventions and diagnoses, highlighting the importance of individualized approaches and further research.
We suggest that the following gaps in effectiveness and implementation research should be addressed in the future to further advance the field:
First, we recommend that future primary studies and reviews should examine and report the impact of lifestyle interventions on both mental and physical health outcomes such as symptom ratings, cognitive and social functioning, cardiovascular risk factors and also quality of life. One of the major strengths of lifestyle interventions is their ability to address several domains, and understanding the holistic benefits of these interventions is crucial. To date, most systematic reviews have focused on body weight as primary outcome. Lifestyle interventions can be beneficial even if they do not result in weight loss, and overemphasising body weight as outcome can demotivate the individuals affected and potentially increase (self-)stigmatisation [Reference Wharton, Lau, Vallis, Sharma, Biertho and Campbell-Scherer136]. Negative attitudes towards individuals with obesity are widespread also among healthcare professionals. Thus, it is crucial to ensure that research does not contribute to the perpetuation of stigma and self-stigma [Reference Wharton, Lau, Vallis, Sharma, Biertho and Campbell-Scherer136–Reference Fond, Vidal, Joseph, Etchecopar-Etchart, Solmi and Yon140].
Second, our meta-review highlighted a relative dearth of evidence relating to interventions in adults with bipolar disorder. Therefore, future research should aim to fill the knowledge gaps in this population.
Third, we noted that physical activity is the best-investigated lifestyle intervention in people with SMI to date. Indeed, even within the latest WHO guidelines on physical activity and sedentary behaviour, schizophrenia and major depression are included among the conditions with sufficient evidence to recommend physical activity in the treatment [Reference Bull, Al-Ansari, Biddle, Borodulin, Buman and Cardon141]. Still, we need to improve our knowledge of the optimal delivery of physical activity interventions. Since there is no clear evidence that one type of exercise is superior to another, but that a minimum of 150 min of at least moderate-intensity physical activity per week is advisable [5], it is crucial to gain further insights on how interventions can be tailored to meet the individual needs of people with SMI and how people with SMI can be optimally supported to exercise regularly. To this end, contextual and motivational factors, such as social support [Reference Cabassa, Camacho, Vélez-Grau and Stefancic142], exercise history and preferences [Reference Tew, Bailey, Beeken, Cooper, Copeland and Brady143, Reference Sherwood and Jeffery144], as well as exercise opportunities, including financial support [Reference Tew, Bailey, Beeken, Cooper, Copeland and Brady143], appear to be important and should be considered in exercise referrals [Reference Diamond, Waite, Boylan, Hicks, Kabir and Freeman145–Reference Tomlinson-Perez, Machaczek, Firth, Pollard, Meda and Keddie147].
Fourth, as previously noted [Reference Firth, Solmi, Wootton, Vancampfort, Schuch and Hoare34], there is a current paucity of trials and reviews on the effects of sleep and dietary interventions and to date, mental illnesses have not been a specific focus within related public health guidelines. Therefore, future high-quality trials are needed to assess the therapeutic potential of these two important areas.
Fifth, clinicians currently lack appropriate tools to comprehensively assess lifestyle factors in standard care. Relying solely on anthropometric measures alone may result in delayed interventions, which could hinder the effective prevention of secondary diseases [Reference Firth, Siddiqi, Koyanagi, Siskind, Rosenbaum and Galletly134]. Behaviours related to diet (dietary patterns), physical activity (accelerometric data or self-report assessment tools, e.g., [Reference Sallis148–Reference Rosenbaum, Morell, Abdel-Baki, Ahmadpanah, Anilkumar and Baie150]), and sleep (sleep–wake cycle and quality assessments) should be considered alongside physical and metabolic assessments for a more accurate health risk evaluation.
Sixth, we need to gain a deeper understanding of the underlying biological, psychological, and social mechanisms through which lifestyle interventions impact mental health [Reference Firth, Solmi, Wootton, Vancampfort, Schuch and Hoare34]. This will enable us to tailor lifestyle interventions to the unique needs of individuals with different SMI and disease phases (and with mental and physical multimorbidity) and to identify specific strategies and approaches that are most effective for different subgroups. Furthermore, we need to understand potential interactions between psychotropic medications, psychotherapy, and lifestyle interventions and how best to use lifestyle interventions in conjunction with established therapies.
Seventh, research should also focus on effective ways to implement lifestyle interventions within mental health treatment settings. This includes integrating these interventions into existing mental health care systems, training mental health professionals in their use, involving experts such as exercise physiologists, physiotherapists or nutritionists in their delivery, and understanding barriers and facilitators to implementation [Reference Firth, Siddiqi, Koyanagi, Siskind, Rosenbaum and Galletly134]. As lifestyle interventions represent a multifaceted treatment approach, we believe it is valuable to build interdisciplinary networks for delivery and to regularly seek input from allied health professionals [Reference Firth, Siddiqi, Koyanagi, Siskind, Rosenbaum and Galletly134]. In addition, technology-based interventions, such as mobile apps or online platforms, might be helpful to increase accessibility and to promote engagement in healthy lifestyle behaviours.
Eighth, lifestyle interventions in general have the potential to be provided at relatively low cost but given the often limited resources in local community settings, future studies should also include economic evaluations to provide convincing data on the economic benefits of lifestyle interventions in adults with SMI.
Finally, we should not only examine the short-term effects of lifestyle interventions, as it is currently most often the case due to feasibility constraints (study funding only for specific durations and lacking observational data on long-term uptake and efficacy), but also explore strategies for sustaining positive lifestyle changes and investigate their long-term impact on physical and mental health outcomes. Overall, addressing these research gaps will contribute to a more comprehensive understanding of how best to use lifestyle interventions.
Results from this guidance paper should be interpreted within its strengths and limitations. The strength of this guidance paper lies in its comprehensive approach covering several domains of the heterogeneous field of lifestyle interventions in SMI populations and its transparent methodology for evidence synthesis and recommendation development. However, summarizing the evidence and developing recommendations highlighted several challenges, arising mainly from the heterogeneity of the available evidence. Since our guidance paper is based on evidence from already existing systematic reviews, we have thus omitted primary studies that have not yet been included in reviews. In addition, we focused on the most common types of interventions related to physical activity, diet and sleep. Although these broad categories cover a wide spectrum of interventions, we left aside evidence from other emerging lifestyle domains, such as those targeting loneliness and social interactions, related to the workplace and working conditions, or stress management and mindfulness-based interventions. In some cases, we had to artificially categorise lifestyle behaviours that span several domains and may have effects at multiple levels within a single category. The interventions within each category also varied considerably, making it difficult to summarise them for the development of recommendations. Moreover, many factors might moderate the beneficial effects of lifestyle interventions, some of which are often under-reported. For example, the type of professional supervision, the size of the group and the individual adaption of the interventions might have a huge impact on results. These adaptations are particularly important to ensure applicability to special patient populations, such as those with physical disability or cognitive impairment. In some cases, lifestyle interventions might be generally unacceptable or inappropriate, so that medication options should be considered for specific outcomes, for example, metformin for antipsychotic-induced weight gain [Reference Fitzgerald, O’Connell, Keating, Hynes, McWilliams and Crowley151]. Future research should also assess these aspects in more detail and implement them in further recommendations. Finally, we focused on lifestyle interventions in adults with SMI. The degree to which the results can be transferred to children and adolescents living with SMI is not clear and requires further study and a separate review.
Conclusions
This guidance paper provides a comprehensive summary of the current meta-review evidence regarding the benefits of lifestyle interventions in adults with SMI and identified current research gaps. Evidence supports the application of lifestyle interventions that combine behavioural change techniques, dietary modification, and physical activity to reduce weight and improve cardiovascular health parameters in adults with SMI. Our meta-review indicated a relative lack of research regarding interventions for adults with bipolar disorder, as well as on nutritional and sleep interventions. With further research, lifestyle interventions can and should be a core component of mental health care in the future.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1192/j.eurpsy.2024.1766.
Author contribution
All authors contributed to the study conception and design of the study. Study methodology was developed by IM, SW, JS, AV and SM who were supervised by AS, AH and PF. IM, SW, JS, AV, SM, VS were responsible for data acquisition, curation and interpretation. IM, SM, JS, and SW wrote the original draft. All authors were involved in revising the article; read and approved the final version of the manuscript.
Financial support
The work was supported by the Else Kröner-Fresenius Foundation to PF, AS, and IM (Residency/PhD track of the International Max Planck Research School for Translational Psychiatry [IMPRS-TP]). This work was endorsed by DZPG (German Center for Mental Health) (to AH, FKZ: 01EE2303C; to PF, FKZ: 01EE2303A, 01EE2303F).
Competing interest
MS received honoraria/has been a consultant for Angelini, Lundbeck. JF is supported by a University of Manchester Presidential Fellowship (P123958) and a UK Research and Innovation Future Leaders Fellowship (MR/T021780/1) and has received honoraria/consultancy fees from Atheneum, ParachuteBH, Richmond Foundation and Nirakara, independent of this work. CC has been a consultant and/or advisor to or has received honoraria from: AbbVie, Acadia, Adock Ingram, Alkermes, Allergan, Angelini, Aristo, Biogen, Boehringer-Ingelheim, Bristol-Meyers Squibb, Cardio Diagnostics, Cerevel, CNX Therapeutics, Compass Pathways, Darnitsa, Denovo, Gedeon Richter, Hikma, Holmusk, IntraCellular Therapies, Jamjoom Pharma, Janssen/J&J, Karuna, LB Pharma, Lundbeck, MedAvante-ProPhase, MedInCell, Merck, Mindpax, Mitsubishi Tanabe Pharma, Mylan, Neurocrine, Neurelis, Newron, Noven, Novo Nordisk, Otsuka, Pharmabrain, PPD Biotech, Recordati, Relmada, Reviva, Rovi, Sage, Seqirus, SK Life Science, Sumitomo Pharma America, Sunovion, Sun Pharma, Supernus, Takeda, Teva, Tolmar, Vertex, and Viatris. He provided expert testimony for Janssen and Otsuka. He served on a Data Safety Monitoring Board for Compass Pathways, Denovo, Lundbeck, Relmada, Reviva, Rovi, Supernus, and Teva. He has received grant support from Janssen and Takeda. He received royalties from UpToDate and is also a stock option holder of Cardio Diagnostics, Kuleon Biosciences, LB Pharma, Mindpax, and Quantic. AS was an honorary speaker for TAD Pharma and Roche and a member of Roche advisory boards. AH was member of advisory boards of Boehringer-Ingelheim, Lundbeck, Janssen, Otsuka, Rovi and Recordati and received paid speakership by these companies as well as by AbbVie and Advanz. He is editor of the German schizophrenia guideline. PF has been EPA president in 2022 and is a co-editor of the German (DGPPN) schizophrenia treatment guidelines and a co-author of the WFSBP schizophrenia treatment guidelines; he is on the advisory boards and receives speaker fees from Janssen, Lundbeck, Otsuka, Servier and Richter. IM, SW, JS, AV, SM, VS, PvP, BS, DV, TK, MH, MG, WG and HJM report no conflicts of interest.
Open access
Comments
No Comments have been published for this article.