Multiple versus single risk behaviour interventions for people with severe mental illness: a network meta-analysis and qualitative synthesis

Model 1: intervention effects model

Across studies, interventions targeted a range of single (e.g. unhealthy diet alone, physical inactivity alone) or multiple health risk behaviours (e.g. unhealthy diet and physical inactivity). We used NMAs to assess the effectiveness of targeting these different combinations of risk behaviours.

Studies varied in terms of participant (e.g. targeting people with SMI who also had comorbid physical illness) and intervention characteristics (e.g. whether studies were primarily delivered to groups or individuals). Therefore, we extended the model to include covariates estimating the association between these participant and intervention characteristics and intervention effectiveness.

Model 2: does targeting multiple risk behaviours lead to positive or negative synergies (interaction model)?

Model 1 gives some insight into the benefits of targeting different combinations of risk behaviours. But it does not directly assess whether or not targeting multiple risk behaviours leads to positive or negative synergies (e.g. does targeting diet and physical activity together result in greater or fewer benefits than expected from the sum of their effects?).

There were sufficient data to conduct interaction models for weight and BMI models only. We used a components NMA approach,37 where µ_j is the mean weight (kg) or BMI (kg/m²) for the TAU group b_i in trial j. θ_jk is mean weight (kg) or BMI for any risk behaviour intervention k from trial j:

where

δ_jk represents the MD between TAU and any risk behaviour intervention k, compared with TAU b_j, in trial j, with between-study SD τ; d_k is the pooled estimate of the MD comparing any risk behaviour intervention with TAU.

Model 2 is an extension of model 1a:1, in which the MD d_k was the MD comparing any intervention with TAU, which assumed that:

Model 2 allowed for each component to have its own independent impact on the MD. For example, targeting diet and physical activity together may each have contributed separately to reductions in weight or BMI. The model also allows components to interact with each other. For example, targeting diet and physical activity may lead to greater reductions (positive synergies) in weight or BMI than would be expected by the sum of these components. Model 2 also allowed for lesser reductions in weight or BMI than would be expected if the components were summed (negative synergies). We did not include targeting alcohol or smoking as separate effects as all interventions in the analyses targeting one of these behaviours also targeted diet and physical activity:

where PA = physical activity.

Model 3: effectiveness of behaviour change techniques

Risk behaviour interventions vary not only in terms of which behaviours are targeted, but also in terms of the BCTs included in these interventions.

We therefore fitted models that classified interventions in terms of BCTs (using the BCTT v1). 33 We fitted two models and compared them in terms of goodness of fit (total residual deviance, between-study SD and DIC). As above, there were sufficient data to conduct these analyses for only weight and BMI.

Smoking

Cannabis

Four trials, with 716 participants providing data, were included in the NMA on cannabis use. 47,49,67,74

Each trial targeted a different combination of risk behaviours: alcohol, drugs or both;49 diet, physical activity, smoking and alcohol;47 physical activity, smoking and alcohol;73 and smoking, diet, physical activity, alcohol, any drug use. 67

Therefore, there were sufficient data to only compare any intervention with TAU at the end point. Given the sparse nature of the network, we decided to use a fixed-effects model, as there were insufficient data to reliably estimate between-study heterogeneity.

Goodness of fit was acceptable for the model (total residual deviance was 3.14 data points, compared with 4 data points). There was no evidence that interventions reduced cannabis use (SMD –0.05, 95% CrI –0.25 to 0.15). However, CrIs were relatively wide, indicating uncertainty about this effect estimate.

Alcohol misuse

Five trials (731 participants) were included in the NMA on alcohol use. 47,49,67,73,78 Trials targeted a range of combinations of risk behaviours: alcohol alone;78 alcohol, drugs or both;49 diet, physical activity, smoking and alcohol use;47 physical activity, smoking and alcohol use;73 and smoking, diet, physical activity, alcohol use and drug use. 67

There were sufficient data to only compare any intervention with TAU at the end point. Goodness of fit was acceptable for the model (total residual deviance was 5.25 data points, compared with 5 data points) and the between-study SD was 0.36 (95% CrI 0.02 to 2.01, SMD scale). There was no evidence that interventions reduced alcohol misuse (SMD 0.14, 95% CrI –0.71 to 1.00). CrIs were very wide, indicating considerable uncertainty about this effect estimate.

Physical activity

Seven trials45,67,86,89,105,109,149 reported data on physical activity. Most of these trials measured total physical activity (including self-report measures, such as the International Physical Activity Questionnaire, and objective measures, such as an accelerometer).

Five trials45,75,86,89,105 targeted diet and physical activity; one trial targeted physical activity and sedentary behaviour;109 and one trial targeted diet, physical activity, smoking, alcohol misuse and drug misuse. 67

Therefore, there were sufficient data to conduct only meta-analyses of any intervention and TAU. Goodness of fit for this model was acceptable [total residual deviance was 8.51 data points, compared with 7 data points and between-study SD was 0.23 (95% CrI 0.01 to 0.96)]. Evidence was very limited for the effect of any interventions compared with TAU for improving total physical activity (SMD 0.10, 95% CrI –0.25 to 0.49).

Diet

Six trials45,47,67,88,89,93 reported data on diet. Diet was measured using a diverse range of outcomes: calorie intake,89 fat intake45,67,93 and fruit and vegetable intake. 47,88 We considered these outcomes too diverse to combine in meta-analyses. Therefore, despite an unhealthy diet being one of the most commonly targeted health risk behaviours, it is difficult to conclude how effective interventions are at promoting a healthier diet.

Anthropometric outcomes

Cardiovascular-related outcomes

Quality of life

Quality-of-life data help to assess the overall self-reported benefits of interventions to physical health. In addition, these data help to assess if there are any self-reported benefits or deterioration to mental health-related quality of life.

Positive and Negative Syndrome Scale

The Positive and Negative Syndrome Scale (PANSS) was the most commonly reported mental health outcome measure; it is designed to monitor symptom severity in people with schizophrenia. We used these data to assess whether or not risk behaviour interventions had adverse effects on mental health for people with SMI.

There were insufficient data to assess the impact of which behaviours were targeted; therefore, we assessed differences between any risk behaviour intervention and TAU on the PANSS. As there were only two comparisons in the network, this was equivalent to conducting a pairwise meta-analysis. In addition, only data on the PANSS total score51,54,60,67,79 were sufficiently reported to conduct meta-analyses.

Given that only five trials were included in the analysis, and concerns about whether or not there were sufficient data to estimate between-study heterogeneity, we compared the goodness of fit between a random-effects and a fixed-effects model. The greater complexity of the random-effects model did not result in a better fit with the data (random-effects model DIC = 56.91, total residual deviance = 10.25 data points; fixed-effects model DIC = 56.63, total residual deviance = 11.09 data points, compared with 10 data points); therefore, we used a fixed-effects model for the analysis of the PANSS total score.

There was no evidence on the PANSS total score (MD 0.03, 95% CrI –2.56 to 2.65) that any risk behaviour intervention had a harmful effect on mental health in people with schizophrenia, although the CrI was wide, suggesting uncertainty on estimating the true effect.

Weight

We investigated the impact of including five study characteristic covariates to NMA model 1a: behaviour change intervention at initiation of antipsychotic treatment, if participants were targeted for specific comorbid conditions (e.g. diabetes), if the intervention was delivered to individuals, if the intervention was delivered in an inpatient setting and if the intervention was tailored for people with SMI. We then compared the goodness of fit of this covariate model with model 1 without covariates.

We compared model 1b (covariate model) with model 1a (model without covariates). Total residual deviance showed that goodness of fit was acceptable in both models (model 1b = 55.30 data points, model 1a = 60.00 data points, compared with 63 data points). The DIC did not substantially differ between models (model 1b = 279.52, model 1a = 282.17).

However, the inclusion of covariates in model 1b explained some of the heterogeneity, indicated by a reduction in the between-study SD (model 1b: SD 0.44, 95% CrI 0.02 to 1.54; model 1a: SD 0.69, 95% CrI 0.04 to 1.78).

Interventions delivered to individuals were more effective than those delivered to groups (beta –2.70, 95% CrI –4.69 to –0.75). In addition, interventions delivered in inpatient settings were less effective than interventions delivered in community settings (beta 3.18, 95% CrI 0.21 to 6.12). CrIs were very wide for all other covariates.

Body mass index

Differences in DIC values indicated that the covariate model (model 1b: DIC = 186.91) fitted the data better than the model without covariates (model 1a: DIC = 196.76), although the covariates included in the model reduced the between-study SD only a little (model 1b: SD 0.83, 95% CrI 0.46 to 1.40; model 1a: SD 0.88, 95% CrI 0.56 to 1.35).

Total residual deviance was acceptable for the covariate model (mean 68.53 data points, compared with 70 data points). Similar to weight outcomes, interventions delivered to individuals were more effective than those delivered to groups (beta –1.11, 95% CrI –2.15 to –0.01). All other covariates had very wide CrIs.

Weight

Model fit was acceptable (mean total residual deviance = 58.42 data points, compared with 63 data points). The findings were similar to those of model 1. We did not find evidence of positive interactions. There did not appear to be additional benefits of targeting diet and physical activity concurrently, compared with summing the effects of targeting each of these behaviours alone (MD 1.88 kg, 95% CrI –4.02 to 7.57 kg).

There was evidence of negative synergistic effects when targeting diet, physical activity, smoking, alcohol misuse and drug misuse. Benefits on weight loss were less than would be expected if these behaviours were targeted alone (2.97 kg, 95% CrI 1.03 to 4.84 kg).

Body mass index

Model fit was acceptable (total residual deviance was 71.78 data points, compared with 75 data points). Targeting physical activity was associated with a small reduction in BMI (MD –1.08 kg/m², 95% CrI –2.28 to 0.13 kg/m²). There was insufficient evidence for the effectiveness of targeting diet alone (MD –0.01 kg/m², 95% CrI –1.65 to 1.67 kg/m²).

There was no evidence of a positive interaction when targeting diet and physical activity concurrently (0.56 kg/m², 95% CrI –1.48 to 2.61 kg/m²). There was also no evidence of a positive interaction when targeting diet, physical activity, smoking and alcohol misuse (MD 0.66 kg/m², 95% CrI –0.43 to 1.68 kg/m²).

Weight

Goodness of fit was acceptable for both models (independent model: total residual deviance = 59.42 data points, interaction model: total residual deviance = 58.46 data points, compared with 63 data points).

The DIC values (independent model: DIC = 281.67, interaction model: DIC = 282.37) and between-study SDs were similar (independent model: SD 0.81, 95% CrI 0.03 to 2.22; interaction model: SD 0.91, 95% CrI 0.04 to 2.43). This may reflect that there were sufficient data for only two interactions to be included in the interaction model and that both covariates were very imprecise. Therefore, the summary of results will focus on the simpler independent model.

The strongest evidence was found for goal-setting (behaviour) (MD –2.22 kg, 95% CrI –4.54 to –0.44 kg) and instruction on how to perform the behaviour (MD –2.10 kg, 95% CrI –3.42 to –0.45 kg).

In addition, problem-solving was associated with weight gain (MD 3.61 kg, 95% CrI 1.18 to 5.32 kg).

Body mass index

Goodness of fit was acceptable for both models (independence model: total residual deviance = 77.86, interaction model: total residual deviance = 73.31).

The interaction model (DIC = 192.86) fitted the data better than the independent model (DIC = 198.84) and explained more of the heterogeneity [between-study SD: interaction model, 0.21 (95% CrI 0.01 to 0.74); independence model, 0.55 (95% CrI 0.12 to 1.07)]. Therefore, the summary will focus on results from the interaction model.

Similar to the weight NMA, the strongest evidence was found for goal-setting (behaviour) (MD –1.85 kg/m², 95% CrI –2.91 to –0.69 kg/m²), instruction on how to perform a behaviour (MD –1.19 kg/m², 95% CrI –1.85 to –0.55 kg/m²) and self-monitoring of behaviour (MD –0.70 kg/m², 95% CrI –1.42 to 0.07 kg/m²), although it was not possible to rule out negligible benefits for self-monitoring of behaviour.

As for weight, problem-solving was associated with a worsening of BMI outcomes (MD 1.46 kg/m², 95% CrI 0.11 to 2.67 kg/m²).

Healthy behaviours improving mental health and well-being

People with SMI reported that physical activity was helpful in changing mental health. But, overall, mental health and a sense of well-being were valued more than changes in activity.

When discussing physical activity, data gathered around two core ideas: physical activity to refresh the mind and physical activity for symptomatic relief. Engaging in activity was frequently reported to result in mental clarity:

. . . I’d notice, after I’d done the gym sessions, I felt really fresh. Like, afterwards, like my mind felt kind of washed, if you like.

Firth et al. ,119 single risk behaviour intervention (SRB), physical activity

Healthy behaviours were also used for immediate symptom relief in one study:

. . . when I feel like I’m just so angry, and stuff like that, I’ve gone training, I’ve done press-ups and stuff at home and it’s cleared my head, killed me anger . . .

Firth et al. ,119 SRB, physical activity

An overarching concept from this theme related to participants’ desire to find a suitable space to bring about positive mental change:

[The gym] gives me somewhere to go to vent out stress.

Jimenez et al. ,127 multiple risk behaviour intervention (MRB), physical activity and diet

Similarly, reports of physical activity to relieve mental health symptoms spoke of exercise as a mental release, and relieving SMI symptoms:

Exercise relieves stress, sweat relieves stress; those two things, I really use them. There’s all kinds of ways you can relieve stress. I think exercise helps me to relieve stress.

Jimenez et al. ,127 MRB, physical activity and diet

The voices, I do not hear them as much.

Forsberg et al. ;121 MRB; physical activity, diet, alcohol, drug use, smoking

There was also widespread recognition that exercise could have immediate impact on well-being, as well as physiological benefits:

. . . physical exercise, it lets endorphins off in your brain, doesn’t it. So even if you are a bit depressed, it can, erm, make you not depressed [laughs], or make you slightly more happy.

Firth et al. ,119 SRB, physical activity

Individual factors

Various personal motivations to change were reported. Identifying and holding on to a personal motivation was noted as an important guiding light:

I think it’s important to, you know, like, have a, like, a north star like that to look at, like, this is why I’m doing it and this is why it’s important.

Heffner et al. ,123 SRB, smoking

The commitment to succeed, in the face of challenges presented by SMI, was also important:

. . . I felt as though I took it serious and I wanted to do it, and then something would come along and sort of like take my mind off everything.

Knowles et al. ,131 SRB, smoking

Motivation stemming from concerns with physical appearance and weight were reported in some studies: ‘I’m a bit fat and have a big stomach.’ (Forsberg et al. ,121 MRB; physical activity, diet, alcohol, drug use, smoking), and they remember being ‘skinny . . . proportioning’ (Jimenez et al. ,127 MRB, physical activity and diet). Concerns over personal appearance could also affect concerns for romantic lives (related to concepts of confidence and acceptance; see Valued benefits of behaviour change interventions):

I want to get married one day or have a girl and girls don’t like fat guys. I want to be able to have a girl and be confident.

Jimenez et al. ,127 MRB; physical activity and diet

This sense of participants wanting to promote a more positive future version of themselves extended to how they looked to protect themselves from the negative health outcomes associated with long-term conditions and ageing:

I will soon be over 50 years old; you must be aware of the blood sugar and such values . . . It can change. It gives the desire to keep good values and not to get diabetes and other diseases.

Rönngren et al. ,140 MRB, physical activity and diet

For others, motivations to change to healthier patterns of behaviour stemmed from an awareness of familial risks and the impact of witnessing others suffer health consequences of weight gain and smoking:

It seems there’s a lot of health issues with family and that plays a part in wanting to exercise and lose weight because I’m at severe risk of getting all these things.

Jimenez et al. ,126 MRB, physical activity and diet

My dad, as he’s got a diseased lung . . . he smokes a lot and just watching him, cos he’s a pack a day smoker, I don’t want to be that.

Thornton et al. ,142 SRB, smoking

In contrast to these direct health concerns, one study reported motivation stemming from particular scenarios, such as avoiding associated risks of possible surgery for weight loss:

I don’t want to have to go through any weight loss surgery and stuff like that. That’s stuff I worry about.

Yarborough et al. ,147 MRB, physical activity and diet

Social factors

Feedback and reinforcement of positive messaging from family and friends was also critical to some participants’ motivations to engage in more positive health behaviours:

‘Gee, you look better’, so I’m on the right road.

Aschbrenner et al. ,116 MRB, physical activity and diet

My family is starting to notice that I’m losing weight. I like the positive comments.

Yarborough et al. ,147 MRB, physical activity and diet

A sense of belonging could also be found in motivation from peers, and group interventions were reported as a good starting point to engaging in exercise:

Exercise in a group could be a good way to start when it comes to engaging in physical activity.

Rönngren et al. ,141 MRB, physical activity and diet

The presence of help suggested increased engagement with support. Having ‘someone to call and talk things over with . . . 24 hours a day every day’ (Maisto et al. ,133 MRB, alcohol and drug use). The idea of the intervention as a constant to return to was also present:

So maybe if you just keep . . . keep going back to them. You’re saying ‘look, we’re here and we’ll keep letting you know we’re here’.

Knowles et al. ,131 SRB, smoking

However, verbal support alone did not necessarily translate into helpful actions (needing to be seen as credible), as, in one study, someone reported:

My husband verbally supports me but he eats whatever he wants to eat and has it in the house.

Yarborough et al. ,147 MRB, physical activity and diet

The presence of key others could positively support behaviour change, but not always, with some studies reporting that goals to enact behaviour change were viewed with scepticism:

[W]hen I was talking about quitting there was [sic] a couple of people that were like would verbally basically be like, ‘Yeah, right, I’ll believe it when I see it.’.

Heffner et al. ,123 SRB, smoking

Accountability to friends, family and intervention groups was a prominent component of this subtheme. Group interventions offered inherent accountability among peers. Interpersonal relationships with family and friends offered the ability to monitor behaviour, which was supportive of positive efforts made by individuals. 133 In one study, participants arranged accountable relationships between themselves, calling each other. 126

Peer support was important for some participants, who derived a sense of accountability to each other by their membership and participation of intervention groups:

The accountability of being part of the group is very important.

Yarborough et al. ,147 MRB, physical activity and diet

This level of accountability to the group often translated into supporting each other to adhere to positive behaviour change:

He’ll ask, ‘Did you go to the gym?’ If I say no, he’ll say, ‘I’m disappointed at you.’. But you see, that’s all we got, is this friendship. We have nothing, so we make the best of nothing.

Jimenez et al. ,126 MRB, physical activity and diet

In another instance, the support of friends was often experienced in direct and instructional ways:

I have this friend, she has this attitude, if I say that I’m not going somewhere she’s like, ‘You’re so going. Get dressed.’. Then I have to go.

Jimenez et al. ,126 MRB, physical activity and diet

Another component of this subtheme was surveillance and scrutiny from others. Familial involvement or surveillance could stir individuals into action, not necessarily for themselves, but to appeal to the desires and wishes of others and preserve good relations with significant others:

. . . my parents told me that this treatment was a gift – the sort of habits and things that I was getting into wasn’t something that my parents could tolerate forever.

Luciano et al. ,132 MRB, alcohol and drug use

In another study, motivation to change stemmed from a desire to not become stigmatised or attract negative attention:

I don’t want to have to deal with people’s negative comments about me smoking because I already feel that in other areas.

Heffner et al. ,123 SRB, smoking

Environmental factors

Having a safe and stable place might be the foundation for people with SMI to consider the challenges of behaviour change:

Safe environments are the most important thing to people with mental illness.

Huck et al. ,125 SRB, physical activity

Therefore, environmental factors refers to the external context of an individual on a micro level, for example where they live, and structures created by, or imposed on, a person.

Creating schedules was an important means to shape and structure environments for individuals. Scheduling also offered accountability and was interpreted as motivating. This included scheduling intervention meetings or even behaviour change activities. For physical activity, this notion of environmental scheduling related to practical elements such as specific times, locations and goals that could be set objectively. By structuring their environments through scheduling, participants were, in some senses, bound to complete tasks linked to behaviour change:

I guess it’s because it’s a little motivator. It’s like I can’t get away with it [skipping the gym].

Aschbrenner et al. ,116 MRB, physical activity and diet

Pacing the frequency of intervention meetings was an important consideration. Meetings may prompt behaviour change, rather than just acting as check-in points. For one participant, if the meetings were too infrequent, then they would engage in risk behaviours and:

. . . a week later [as the monthly meeting approaches] I’ll try to improve on it.

Yarborough et al. ,147 MRB, physical activity and diet

A feature of environmental factors contributing to motivation to change were interactions with formal health care. Appointments with intervention providers were another form by which environmental scheduling drove engagement with positive health behaviours. Here ‘providers’ refers to those providing, organising and/or delivering the intervention, whether medical professional or community organiser. The authority held by a provider could increase the sense of accountability felt by participants to stick to their schedules. One participant stated:

You see, when I got an appointment or something I’m very faithful to that. Like he tells me, ‘Come tomorrow’, I got it in my mind I gotta come to him tomorrow. So I won’t miss that appointment.

Jimenez et al. ,126 MRB, physical activity and diet

Providers could set the tone and make participants feel welcome and bring a sense of humanity:

I enjoyed the warmth of the welcome I got from the staff and the interest they showed.

Roberts and Bailey,139 MRB, other – weight

My therapist, even though I’m a mentally ill psychiatric patient in a mental institution, she still treats me like everybody else.

Maisto et al. ,133 MRB, alcohol and drug use

In one instance, a provider was described as confrontational, but also as a source of stability to promote change:

She used to confront me on it like that . . . she didn’t push me away . . . it helped me quit.

Maisto et al. ,133 MRB, alcohol and drug use

Positive relationships with providers were interpreted as supportive. Providers could be personal cheerleaders or provide a touchpoint for change:

He was like persuading us on like ‘come on you can do it, you can do it!’ when he says things like that it makes me more activate.

Firth et al. ,119 SRB, physical activity

[O]nce [the exercise physiologist and nursing staff] helped me out I began to get fitter, not a great deal fitter, just enough to get through the day.

Fogarty and Happell,120 SRB, physical activity

A positive relationship could also benefit perceptions of intervention providers, as one participant was unaware of their positive impact:

I didn’t know they could do some good things, some of them, I didn’t really know that.

Forsberg et al. ;121 MRB; physical activity, diet, alcohol, drug use, smoking

Individual factors

One of the main contributors to individual barriers to change was SMI symptoms. Across studies, SMI symptoms presented challenges to engaging in behaviour change interventions, as exemplified in participants’ accounts about difficulties with concentrating and difficulties accepting information when delusional:

Many of us can get tired and lose concentration . . . When you’ve had a psychosis you work a lot with thoughts darting around inside your head.

Wärdig et al. ,144 MRB; physical activity, diet, alcohol use and smoking

[I] probably just think it’s a load of rubbish or, you know . . . Because like when you’re delusional you don’t, well I know that . . . I don’t believe anything.

Thornton et al. ,142 SRB, smoking

Deteriorations in mental health were often emphasised by participants as their main concern; therefore, enacting positive behaviour change was of secondary importance:

It’s hard to make changes in your diet and follow the routine . . . when you’re at a point where you just don’t care.

Yarborough et al. ,147 MRB, physical activity and diet

Mental health treatments were also identified as restricting the scope to change behaviours through their impact on bodily systems:

When I am feeling down or depressed, I have almost no energy. In these periods, I have to take more psychotropic drugs, resulting in even less energy.

Verhaeghe et al. ,143 MRB, other – health promotion

Poorer mental health on the part of participants had the capacity to magnify barriers that might have ordinarily been surmountable:

My husband drives me there [to the gym] even when I don’t want to be there. I can walk. It’s not that far from my house to the gym [. . .].

Aschbrenner et al. ,116 MRB, physical activity and diet

However, reluctance to engage in physical activity was not always associated with mental health status:

I just get lazy sometimes. I just don’t want to do it, so I sit on the couch and watch TV.

Yarborough et al. ,147 MRB, physical activity and diet

Another feature of individual barriers to behaviour change was derived from participants’ responses about their self-efficacy to enact change. A lack of knowledge or skill affected participants’ capability to carry out healthy behaviours. For example, not understanding how to use exercise equipment as part of a gym routine was cited as a block to effectively engaging in physical activity:

[Knowing] how many repetitions and how to use the handles and set the chair and everything, that would’ve made it a lot more enjoyable for me.

Huck et al. ,125 SRB, physical activity

In the case of taking exercise in public, stigma around body image was prohibitive of activity:

I hate getting into that pool because I think people are staring at me . . . but it’s that whole confidence thing . . . you see those skinny people . . . that’s what helps to put you off . . .

Roberts and Bailey,139 MRB, other – weight

Physiological as well as psychological barriers emerged as common factors in this theme about individual-level barriers to behaviour change. Cravings for addictive substances were commonly reported among participants with histories of addictive behaviours such as smoking, alcohol use and drug use. Cravings appeared to be powerful and intangible, described as ‘screaming’ (Thornton et al. ,142 SRB, smoking) and like a ‘forcefield’ (Maisto et al. ,133 MRB, alcohol and drug use) that could trap people in cycles of negative behaviours.

An element of timing was also present, as cravings were reported at the beginning of a transition:

The first week I felt like every cell in my body wanted a cigarette.

Heffner et al. ,123 SRB, smoking

Distractions were used in response to a craving to engage in a risk behaviour, and building a resilience over time was described:

Maybe call a sponsor, call a friend, go for a walk, and go to the gym. The cravings do pass, and they do pass more easily when you have longer-term sobriety under your belt.

Luciano et al. ,132 MRB, alcohol and drug use

This progression was challenged in one report, as overcoming cravings was described as ‘almost impossible’,142 even when presented with visual consequences of smoking:

[E]ven with those pictures sitting right in front of me because my body is asking me for that.

Thornton et al. ,142 SRB, smoking

Social factors

Influences of the social and cultural contexts were a rich subtheme of data. People with SMI reported the benefits of positive peer support, but also the difficulties with negative peer group pressure.

Participants typically remarked that obligations to family and the cultural significance of dietary behaviours could prevent behaviour change. Eating together was seen as part of their ‘identity’ in a study of a Latino subpopulation:

. . . being together, eating our traditional food with our family, and remembering.

Jimenez et al. ,126 MRB, physical activity and diet

Such family gatherings were an example of social challenges to healthy diet and alcohol consumption:

‘One plate isn’t going to hurt you’, but the plate is like this [gestures] big, and it’s all mounded, it’s huge, and then everyone had to have a drink, and I’m like, ‘I don’t drink anymore’.

Aschbrenner et al. ,116 MRB, physical activity and diet

Engagement in risk behaviours (such as smoking) was often encouraged rather than restrained by others:

People are just like, ‘Dude, if you need to smoke, smoke.’.

Heffner et al. ,123 SRB, smoking

The absence of social support was also detrimental to healthy behaviours. In a few studies, there were reports of the negative impacts of loneliness:

The biggest culprit is loneliness, a terrible problem, I think. Nobody feels good when they’re alone.

Rönngren et al. ,141 MRB, physical activity and diet

Interventions that built in behaviour change activities into people’s schedule interrupted the sense of loneliness felt by some:

[I]t just helps to get me out really, that’s what it does, for people that are isolated like me, that do suffer from mental illness and do find it hard to go out and about.

Hodgson et al. ,124 SRB, physical activity

Environmental factors

The subtheme of environmental barriers to behaviour change emerged from reports about how certain environments induced rather than prevented health risk behaviours.

In some situations, environments were sources of behavioural triggers that were difficult to avoid and promoted engagement in risk behaviours:

You’re sitting there watching it and it’ll come on and triggers cigarette thoughts and . . . instead of putting you off, it would make you want to have a cigarette  . . .

Thornton et al. ,142 SRB, smoking

I’d been smoking for so many years that there’s so many situations where it was extremely automatic for me to smoke.

Heffner et al. ,123 SRB, smoking

In one study, there were reports of individuals seeking new residences to change environmental triggers around them. However, group homes could also be a trigger, as occupants’ shared experiences made for talking points with unintended side effects:

[A]ll those war stories used to get me excited about [using alcohol and drugs].

Maisto et al. ,133 MRB, alcohol and drug use

Another component of environmental barriers to behaviour change was the interface with formal health care. Intervention providers and prescribers are gatekeepers for interventions, with the power to bestow decisions on who gets access to opportunities and support to promote positive behaviour change. This was illustrated through reports of the difficulties participants faced in being prescribed antismoking treatment, as well as in collecting prescriptions:

‘We’re not putting you on the Champix’ [Champix^®; Pfizer Inc., New York, NY, USA] and the other one as well, ‘Not putting you on them’. And that was it. I was out the door, gone.

Knowles et al. ,131 SRB, smoking

Providers may be considered a natural authority in an intervention, which lends their advice legitimacy, as providing advice suggests that there is improvement to be made:

When she [nurse] gives me advice how to cook using healthy ingredients or to go to the fitness, I would follow this advice. The fact that she mentions it means for me that here is a problem.

Verhaeghe et al. ,143 MRB, other – health promotion

However, the naturally authoritative role of providers was sometimes perceived by participants as obstructive. This was especially true when participants perceived that providers lacked knowledge about their lived experience and were guilty of poorly expressing instructions:

How do they know it does that? You ever took drugs?

Maisto et al. ,133 MRB, alcohol and drug use

[H]e [therapist] says ‘that’s because you eat too much and don’t get enough exercise’. That’s pretty blunt, isn’t it? If you’re overweight, the last thing you want someone to tell you is that you’re fat.

Roberts and Bailey,139 MRB, other – weight

Support from health-care providers could also be construed as too sparse and restricted to addressing risk behaviours in isolation from other issues, thereby limiting the relevance of advice:

[O]nly really advice they give is, one time they said, ‘Now really try this time, [patient name]. Really try’. And I was like ‘alright’. They just signed the prescription and I take it away and try, so that was that.

Knowles et al. ,131 SRB, smoking

In a rehabilitation centre, a typically intense setting, care was not perceived as being supportive enough:

All they did was throw books and papers at you and told you to do this and talk about health, health food, and nothing about yourself.

Maisto et al. ,133 MRB, alcohol and drug use

Access to appropriate resources and practical barriers were other features of environmental barriers. The cost of healthy food was raised as a barrier to changing dietary behaviours. This may be common among a SMI population on a fixed income:

If you buy fresh, raw vegetables, they’re expensive. And I’m on a fixed income.

Jimenez et al. ,126 MRB, physical activity and diet

Locating suitable spaces for physical activity was also a common problem that affected participants’ ability to regularly and effectively take part in physical activity:

When we’re at [nursing facility], you could go on a long walk, but it’s all the way around the building and then you can’t really stop anywhere to rest.

Huck et al. ,125 SRB, physical activity

Barriers can mount up and environmental constraints can be multiplicative, such as when bad weather impeded taking outdoor exercise among those with a relatively low tolerance for exercise:

. . . I was out walking and focusing on my diet. And then I got sloppy with that. I don’t like to walk in the mall, and I don’t like to walk in the cold, the rain.

Yarborough et al. ,147 MRB, physical activity and diet

Tailoring interventions to severe mental illness populations

The most prominent intervention consideration reported by people with SMI was the role of mental health and SMI symptoms in behaviour change, as described in the subtheme of individual barriers to behaviour change. Symptoms of paranoia and delusion were highlighted as challenging for receiving information in one study that investigated perceptions of a population health intervention, for antismoking, in people with SMI. 142 This was the only study to consider the impacts of a population health intervention in this group. People with mental health problems are just as willing to bring about positive behaviour change as the general population. However, for some people with SMI, assistance in organising and structuring their lives to help them engage in positive behaviour change was seen as a vital first step:

[Talking in third person] because it’s not that they can’t think for themselves in the same way as other people, but it’s more a case of they can’t organise themselves or their thoughts in the same way as other people.

Knowles et al. ,131 SRB, smoking

In this sense, the structure and organisation of interventions themselves were positively associated with a greater sense of order and regularity on the part of participants. Engaging with an intervention created a basic schedule for participants:

[G]roups run regularly . . . there’s a schedule every week.

Maisto et al. ,133 MRB, alcohol and drug use

The structure provided by this was touched on in different ways, and described as helping to provide order to life more generally:

It means, how shall I say it? Maybe a bit more order in my life. You see, I have Tuesday and Thursday and I know that time I will . . . That will be, yes . . .

Forsberg et al. ;121 MRB; physical activity, diet, alcohol, drug use, smoking

A feature of structure that was present in some interventions was record keeping, particularly for diet and physical activity, but also applied to other behaviours. Having this organisational reference point seemed to link with self-education about personal health:

Well, it wasn’t until I started writing things down that I realized my intake was terrible.

Yarborough et al. ,147 MRB, physical activity and diet

Participants suggested that the routine and structure gained from interventions were important to continuing progress beyond the intervention:

Some of the most helpful things, it’s a combo, the structure and the day-to-day routine that I developed there. It has helped me for when I’m on my own.

Luciano et al. ,132 MRB, alcohol and drug use

Such organisation was also interpreted as applicable to participants’ cognitive approach:

[B]y maintaining an active, healthy lifestyle, helps me take the reins of my life [. . .].

Jimenez et al. ,127 MRB, physical activity and diet

Knowledge and skills to enact behaviour change

Psychoeducation and knowledge provision components were common among interventions. In studies, individuals reported awareness of the consequences of risk behaviours:

As [service provider] has told me, if I don’t start [exercising], COPD [chronic obstructive pulmonary disease], and that’s all I got to say.

Huck et al. ,125 SRB, physical activity

However, people with SMI may be less informed about behaviour change strategies and healthy choices. Some participants found education necessary, and being taught skills to practice healthy behaviours was enabling and enhanced their self-efficacy. While in a physical activity intervention, fear of incorrectly implementing change as a result of a lack of knowledge was prohibitive to taking action:

SMART [self-management and recovery training] gave me the information I needed no matter what I did.

Penn et al. ,138 SRB, alcohol

There are different ways to walk. It was a big step because I have a lack of knowledge of what is good.

Rönngren et al. ,140 MRB, physical activity and diet

The way in which knowledge was imparted by providers was also crucial to how participants responded to advice and encouragement. Non-tailored communication was especially obstructive and led to disengagement from positive behaviour change:

Sometimes I don’t understand [my therapist]. I let her say what she’s got to say, and I just stay quiet.

Maisto et al. ,133 MRB, alcohol and drug use

In other instances, participants recalled that providers had withheld information and not prepared them for what was to come, leading to only partial engagement in activities that they would have preferred to excel at:

I wish I had been pushed more, it would have been good if I could have been on Survivor on TV.

Wärdig et al. ,144 MRB; physical activity, diet, alcohol use and smoking

Physical activity was often referred to as a work in progress, that led to incremental change over time:

We did some small things and then we built up . . . we just walked around the back alley way quite a few times . . . flat ground so we didn’t have to walk up hills . . .

Fogarty and Happell,120 SRB, physical activity

Being prepared to experience physical exertion was cited as a necessary precursor to taking exercise, as individuals who were not comfortable might disengage before starting. Other paths to exercise were reliant on taking personal responsibility for following advice:

The program works if you follow the steps and the things you do. You get what you put into it I guess.

Luciano et al. ,132 MRB, alcohol and drug use

Intentions to enact behaviour change were also prompted in response to receipt of educational interventions that focused on the poor health consequences of negative health behaviours:

I learned a lot of stuff from the group . . . like how . . . beer and alcohol . . . [do] to your body . . . I’m going to be more careful . . .

Maisto et al. ,133 MRB, alcohol and drug use

Finding out that I’m a borderline diabetic . . . about six months ago . . . was also a push too.

Yarborough et al. ,147 MRB, physical activity and diet

A counterpoint to this sense of personal responsibility to change was noted in responses to public health antismoking campaigns, whereby shocking facts and imagery did not prohibit smoking; instead, people interpreted the risks as possible outcomes that were unlikely to affect them:

One’s [antismoking campaign] sitting right in front of me and I’m still smoking a cigarette, but yeah there is a lot of campaigns to scare people from smoking.

Thornton et al. ,142 SRB, smoking

For others, the risk of poorer health outcomes was divorced from a sense of responsibility for their own health; instead, they perceived the likelihood of being unwell as being inevitable:

[S]ometimes I think it’s in my DNA that . . . everything has been harder for me.

Heffner et al. ,123 SRB, smoking

Group delivery formats

A subtheme emerged about the merits of group delivery of interventions, and is related to broader themes about social motivations for behaviour change. This subtheme particularly highlighted the importance of how group formats fostered peer support and engendered a sense of togetherness:

I think a big part, if you’re going to deal with this, is to have people in the same boat with you. That helped – I didn’t know anyone else that would have been a dual diagnosis.

Luciano et al. ,132 MRB, alcohol and drug use

From bonding and togetherness came trust and understanding among participants and signalled how group delivery could also facilitate more tailored delivery to meet the needs of people with SMI. For example, in one study, it was reported that one individual experienced panic attacks during sessions. Peers understood their reactions and knew ‘to leave me alone in that situation or come over and have a quick word or maybe give me a quick nudge to get back on the pitch, and I just think it’s cus they know [pause] about your needs’ (Hodgson et al. ,124 SRB, physical activity). Put simply, peers ‘know where you’re coming from’ (Maisto et al. ,133 MRB, alcohol and drug use).

Another feature of group-based interventions was the point of comparison they provided. Within a group of peers experiencing the same intervention, there is an inherent comparison:

[M]eeting once a week and comparing notes with people, and knowing that I’m not alone with all these obstacles, and that there are people that care.

Yarborough et al. ,147 MRB, physical activity and diet

This could be a negative, as peers may also police engagement in interventions, as reported in a lifestyle intervention:

Half of them don’t listen. Many of them didn’t even wear that pedometer, and I always wore mine. Even when I slept I wore it. I think I was the most diligent one of us all.

Wärdig et al. ,144 MRB; physical activity, diet, alcohol use and smoking

Choices and adapting interventions to suit the individual

In the context of group interventions, it was apparent that, for some participants, the chance to draw up personalised approaches to behaviour change was an important driver for success. This was especially apparent among those taking part in physical activity interventions:

It was good that we got to choose something out of like a list. We could have done all different things . . . I thought the options were quite good. But I would have like to have done boxercise as well.

Firth et al. ,119 SRB, physical activity

Forewarning about the nature of interventions was also highlighted as critical to making informed choices about taking part in interventions:

[P]eople turn up and then it’s not right for them . . . if they’d had the information . . . they wouldn’t have come, well waste their time really . . .

Roberts and Bailey,139 MRB, other – weight

Because individuals favoured interventions that offered choice, they adapted activities to suit their needs and capabilities, for example incorporating physical activity into their daily routine:

This [exercise] is a really challenging thing for me. What I do is doing errands. I walk from place to place rather than take the bus and the train.

Jimenez et al. ,126 MRB, physical activity and diet

Providers can help facilitate the tailoring of activities so that they are suitable for each individual:

The health co-ordinators were great people who made everything feel spontaneous. They were good at finding the right level for each and every one.

Wärdig et al. ,144 MRB; physical activity, diet, alcohol use and smoking

There were instances, however, when interventions had competing priorities owing to the focus on multiple risk behaviours; the delivery of such interventions did not always map to individual needs and preferences:

[T]he walking [is helpful]. The exercise portion of the group, that can get set on the back burner. And I don’t think it should be.

Yarborough et al. ,147 MRB, physical activity and diet

Purpose to activity

Among studies, a subtheme on intervention activities providing a sense of purpose emerged:

It’s just that I’ve got an activity for the afternoon that I’m not sat watching TV, something like that.

Carless and Douglas,118 SRB, physical activity

Similarly, in two studies, interventions were seen as providing something to look forward to:

[G]roup gives you a place to be, a reason to be there.

Maisto et al. ,133 MRB, alcohol and drug use

It’s a routine you get used to I think. It’s like football, I get excited about football on a Monday.

Carless and Douglas,118 SRB, physical activity

More broadly, the purpose of activity also related to the concepts of loneliness (see Barriers to behaviour change, Social factors). Purposeful activity could help resolve loneliness.

Valued benefits of behaviour change interventions

Across studies, the benefits of behaviour change interventions extended beyond health outcomes, with participants reporting improvements in well-being and ability to navigate day-to-day life. Interventions featuring physical activity were particularly prominent in this subtheme.

Cognitive benefits of behaviour change interventions included memory and decision-making:

My memory has returned.

Rönngren et al. ,140 MRB, physical activity and diet

I wouldn’t make any decisions, I was too worried about making the wrong one. I was very scared [pause] of stepping on people’s feet and just everything like that, . . . now I just get on with it.

Hodgson et al. ,124 SRB, physical activity

Valued benefits of intervention experiences also included opening doors for individuals’ everyday functioning and socialising, with greater control over their lives. As interventions were structured, this, in turn, gave individuals a sense of order over their lives. It can be through simple routine to daily life, or returning to education, work and hobbies:

I’m going to school, I’m looking for work, I go home every other week to see my family. I run, I’ve really gotten into that. I have friends now, I go fishing twice a week . . .

Luciano et al. ,132 MRB, alcohol and drug use

[B]y maintaining an active, healthy lifestyle, helps me take the reins of my life and have a connection with the outside and inside world.

Jimenez et al. ,127 MRB, physical activity and diet

Some of the most generally affecting and prevalent benefits reported in the data were improved confidence, self-esteem and optimism. Confidence was described as part of a package of cognitive and emotional changes:

I think I’m in a much better mood . . . I am up, I’m confident, I’m sleeping well, my appetite is good, my self-esteem is healthy . . . I have a brighter outlook, you know.

Heffner et al. ,123 SRB, smoking

SMART [self-management and recovery training – the intervention] gave me pride. It showed me how to get self-worth. It was basically building me up in order to be receptive to everything else. It started with me.

Penn et al. ,138 SRB, alcohol

Participants’ expressions of excitement and fulfilment in achieving change could bring about greater confidence that, in turn, acted as a spur to continue to engage with interventions:

I love it. It’s just like, thinking yeah, I couldn’t do this 4 months ago, now look at me doing it comfortably. Let’s throw it up a gear!

Firth et al. ,119 SRB, physical activity

This positive change, was summed up in feeling the beginning of a fresh start:

I feel like this program has benefited me and I feel like I’ve got my second wind out of this program.

Fogarty and Happell,120 SRB, physical activity

Interaction of mental and physical health

Participants reported that engaging in health behaviours improved their mental health and well-being, and mental health changes affected their engagement with healthy behaviours. Participants in the included studies wanted to be treated holistically, and sometimes felt that interventions had an impersonal focus on behaviours rather than people. There were also data reporting the use of smoking, alcohol use and drug use to help manage mental health symptoms.

Motivational contexts for behaviour change

A range of individual factors were reported to be personally motivating. These included managing health, improving physical appearance and working towards a positive future. Family and friends were also an important source of motivation, providing feedback and reinforcement of positive change. However, others could also be discouraging when they were careless with their words or actions. Interventions were interpreted as a safe and stable place to work on changing their unhealthy behaviours, also providing a schedule and accountability, which helped to make changes.

Barriers to behaviour change

Mental health symptoms affected individuals’ ability to change. For example, it can be difficult to engage in physical activity when feeling a lack of motivation and low mood related to mental health state, likewise with difficulty concentrating when receiving information related to health. However, mental health symptoms were not the only factors affecting motivations. A general lack of motivation to change unhealthy behaviours was also unrelated to mental health for some. Stigma about body image was another factor considered to act as a barrier to change. Social support from family and friends, although often a positive influence, could also act as a barrier. For example, some family members encouraged unhealthy behaviours. Alternatively, a lack of social support was a barrier for persevering with behaviour change. Environmental factors could act as triggers for unhealthy behaviours (e.g. advertisements on television), as could living in group homes where peers shared their experiences of engaging in risk behaviours.

Experiences of behaviour change interventions

It was suggested that people with SMI needed additional help providing structure and assistance in organising their lives. This was considered an important aspect of tailoring interventions for people with SMI. Interventions providing education and imparting skills to promote healthy behaviours were also considered beneficial. Group interventions were helpful for building peer support and also as a point of comparison with others, which could be double-edged if used to police behaviour. Interventions that took into account the needs, abilities and preferences of participants were felt to be beneficial. A wealth of data reported that interventions had an impact on individuals’ lives beyond just improving their health-related behaviours. For example, interventions sometimes provided a structure and purpose to everyday life. Interventions also helped to build confidence: if positive changes were made in one area of their lives, then it was thought possible in other aspects of life.

Search strategy

#1 MeSH descriptor: [Schizophrenia] explode all trees

#2 MeSH descriptor: [Bipolar Disorder] explode all trees

#3 MeSH descriptor: [Psychotic Disorders] explode all trees

#4 MeSH descriptor: [Affective Disorders, Psychotic] explode all trees

#5 ((acute or chronic* or persistent* or serious* or severe*) NEAR/2 (mental* or psych*) NEAR/2 (disease* or disorder* or disturbanc* or ill*)):ti,ab,kw (Word variations have been searched)

#6 (SMI NEAR/3 (individual* or inpatient* or in-patient* or outpatient* or out-patient* or patient* or people or person* or population*)):ti,ab,kw (Word variations have been searched)

#7 ((mental health or psychiatric) NEAR/2 (inpatient* or in-patient*)):ti,ab,kw (Word variations have been searched)

#8 (schizophren*):ti,ab,kw (Word variations have been searched)

#9 (psychotic or psychosis or psychoses):ti,ab,kw (Word variations have been searched)

#10 (bipolar):ti,ab,kw (Word variations have been searched)

#11 #1 OR #2 OR #3 OR #4 OR #5 OR #6 OR #7 OR #8 OR #9 OR #10

#12 (healthy NEAR/2 (diet* or eating)):ti,ab,kw OR (fruit* NEAR/2 (intake or consum* or increase or portion* or serving* or frequenc* or number* or preference* or choice*)):ti,ab,kw OR (vegetablet* NEAR/2 (intake or consum* or increase or portion* or serving* or frequenc* or number* or preference* or choice*)):ti,ab,kw OR (“5 a day”):ti,ab,kw OR (“five a day”):ti,ab,kw (Word variations have been searched)

#13 MeSH descriptor: [Feeding Behavior] explode all trees

#14 MeSH descriptor: [Food Preferences] explode all trees

#15 MeSH descriptor: [Nutrition Therapy] explode all trees

#16 MeSH descriptor: [Diet Therapy] explode all trees

#17 MeSH descriptor: [Diet] explode all trees

#18 MeSH descriptor: [Obesity] explode all trees

#19 MeSH descriptor: [Overweight] explode all trees

#20 #12 or #13 or #14 or #15 or #16 or #17 or #18 or #19

#21 #11 and #20

#22 (physical NEAR/3 (fit* or train* or activ* or endur*)):ti,ab,kw OR (exercis* NEAR/3 (fit* or train* or activ* or endur*)):ti,ab,kw OR ((promot* or uptak* or encourag* or increas* or start* or adher*) NEAR/3 (exercis* or gym* or sport* or fitness)):ti,ab,kw OR ((decreas* or reduc* or discourag*) NEAR/3 (sedentary or deskbound)):ti,ab,kw OR (sport* or walk* or running or jogging or bicycling or biking or swimming):ti,ab,kw (Word variations have been searched)

#23 (active NEAR/1 (travel* or transport* or commut*)):ti,ab,kw (Word variations have been searched)

#24 MeSH descriptor: [Physical Fitness] explode all trees

#25 MeSH descriptor: [Recreation] explode all trees

#26 MeSH descriptor: [Exercise Therapy] explode all trees

#27 MeSH descriptor: [Exercise] explode all trees

#28 MeSH descriptor: [Running] explode all trees

#29 MeSH descriptor: [Jogging] explode all trees

#30 MeSH descriptor: [Swimming] explode all trees

#31 MeSH descriptor: [Walking] explode all trees

#32 #22 or #23 or #24 or #25 or #26 or #27 or #28 or #29 or #30 or #31

#33 #11 and #32

#34 MeSH descriptor: [Smoke] explode all trees

#35 MeSH descriptor: [Smoking Cessation] explode all trees

#36 MeSH descriptor: [Smoking Reduction] explode all trees

#37 MeSH descriptor: [Nicotine] explode all trees

#38 MeSH descriptor: [Tobacco] explode all trees

#39 MeSH descriptor: [Tobacco Smoking] explode all trees

#40 MeSH descriptor: [Tobacco Use Disorder] explode all trees

#41 MeSH descriptor: [Tobacco Use Cessation] explode all trees

#42 MeSH descriptor: [Tobacco, Smokeless] explode all trees

#43 MeSH descriptor: [Tobacco Smoke Pollution] explode all trees

#44 MeSH descriptor: [Electronic Nicotine Delivery Systems] explode all trees

#45 (smoking or antismoking or anti-smoking):ti,ab,kw OR (smoker or smokers):ti,ab,kw OR (“electronic cigarette*” or e-cigarette* or e-cig or e-cigs):ti,ab,kw OR (vape or vapes or vaper or vapers or vaping):ti,ab,kw (Word variations have been searched)

#46 #34 or #35 or #36 or #37 or #38 or #39 or #40 or #41 or #42 or #43 or #44 or #45

#47 #11 and #46

#48 MeSH descriptor: [Alcohol Drinking] explode all trees

#49 MeSH descriptor: [Alcoholic Intoxication] explode all trees

#50 MeSH descriptor: [Alcoholic Beverages] explode all trees

#51 MeSH descriptor: [Drinking Behavior] explode all trees

#52 (beer* or wine* or cider* or alcopop* or spirit or spirits):ti,ab,kw OR (alcohol*):ti,ab,kw OR (drink* NEAR/2 (binge or excessive* or harm* or heavily or heavy or misus* or abus* or consum*)):ti,ab,kw OR (intoxicat* or inebriat* or drunk*):ti,ab,kw (Word variations have been searched)

#53 #48 or #49 or #50 or #51 or #52

#54 #11 and #53

#55 MeSH descriptor: [Substance-Related Disorders] explode all trees

#56 MeSH descriptor: [Inhalant Abuse] explode all trees

#57 MeSH descriptor: [Marijuana Abuse] explode all trees

#58 MeSH descriptor: [Substance Abuse, Intravenous] explode all trees

#59 MeSH descriptor: [Drug Users] explode all trees

#60 ((drug* or substance*) NEAR/2 (use* or using or abuse* or abusing or misuse* or misusing)):ti,ab,kw OR ((heroin or opiate* or cocaine or crack) NEAR/2 (use* or using or abuse* or abusing or misuse* or misusing)):ti,ab,kw OR ((cannabis or marijuana) NEAR/2 (use* or using or abuse* or abusing or misuse* or misusing)):ti,ab,kw OR ((benzodiazepine* or amphetamine* or methamphetamine* or MDMA or ecstasy) NEAR/2 (use* or using or abuse* or abusing or misuse* or misusing)):ti,ab,kw OR (solvent* NEAR/2 (use* or using or abuse* or abusing or misuse* or misusing)):ti,ab,kw (Word variations have been searched)

#61 (“street drug*” or “recreational drug*” or “illicit drug*”):ti,ab,kw OR ((“prescri* drug*”) NEAR/2 (use* or using or abuse* or abusing or misuse* or misusing)):ti,ab,kw OR (“polydrug use*”):ti,ab,kw OR (“inject* drug use*”):ti,ab,kw OR ((needle* OR syringe*) NEAR/3 shar*):ti,ab,kw (Word variations have been searched)

#62 #55 or #56 or #57 or #58 or #59 or #60 or #61

#63 #11 and #62

#64 #21 or #33 or #47 or #54 or #63

#65 (lifestyle NEAR/2 (intervention* or program*)):ti,ab,kw OR (“life style” NEAR/2 (intervention* or program*)):ti,ab,kw OR (“behavior* change” NEAR/1 (intervention* or program*)):ti,ab,kw OR (“behaviour* change” NEAR/1 (intervention* or program*)):ti,ab,kw OR (“risk factor*” NEAR/2 (program* or intervention*)):ti,ab,kw (Word variations have been searched)

#66 (“multifactorial lifestyle” NEAR/1 (intervention* or program*)):ti,ab,kw OR (“health behavior*” NEAR/1 (program* or intervention*)):ti,ab,kw OR (“health behaviour*” NEAR/1 (program* or intervention*)):ti,ab,kw OR (“health risk behavior*” NEAR/1 (program* or intervention*)):ti,ab,kw OR (“health risk behaviour*” NEAR/1 (program* or intervention*)):ti,ab,kw (Word variations have been searched)

#67 (“risk behavior*” NEAR/1 (program* or intervention*)):ti,ab,kw OR (“risk behaviour*” NEAR/1 (program* or intervention*)):ti,ab,kw OR (“health behavior* change intervention*”):ti,ab,kw OR (“health behaviour* change intervention*”):ti,ab,kw OR (“behavior* risk factor* intervention*”):ti,ab,kw (Word variations have been searched)

#68 (“risk behavior*” NEAR/1 (program* or intervention*)):ti,ab,kw OR (“risk behaviour*” NEAR/1 (program* or intervention*)):ti,ab,kw OR (“health behavior* change intervention*”):ti,ab,kw OR (“health behaviour* change intervention*”):ti,ab,kw OR (“behavior* risk factor* intervention*”):ti,ab,kw (Word variations have been searched)

#69 (“behaviour* risk factor* intervention*”):ti,ab,kw OR (“behavior* risk factor* program*”):ti,ab,kw OR (“behaviour* risk factor* program*”):ti,ab,kw OR (“risk behaviour* intervention*”):ti,ab,kw OR (“risk behavior* intervention*”):ti,ab,kw (Word variations have been searched)

#70 (“risk behaviour* program*”):ti,ab,kw OR (“risk behavior* program*”):ti,ab,kw (Word variations have been searched)

#71 #65 or #66 or #67 or #68 or #69 or #70

#72 #11 AND #71

#73 #64 OR #72.

Search strategy

1. exp schizophrenia/ (165,050)

2. exp bipolar disorder/ (55,219)

3. exp psychosis/ (253,831)

4. ((acute or chronic$ or persistent$ or serious$ or severe$) adj2 (mental$ or psych$) adj2 (disease$ or disorder$ or disturbanc$ or ill$)).ti,ab,kw. (18,709)

5. (SMI adj3 (individual$ or inpatient$ or in-patient$ or outpatient$ or out-patient$ or patient$ or people or person$ or population$)).ti,ab,kw. (1481)

6. ((mental health or psychiatric) adj2 (inpatient$ or in-patient$)).ti,ab,kw. (12,391)

7. schizophren$.ti,ab,kw. (151,316)

8. (psychotic or psychosis or psychoses).ti,ab,kw. (86,293)

9. bipolar.ti,ab,kw. (82,270)

10. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 (373,041)

11. (healthy adj2 (diet$ or eating)).ti,ab,kw. (15,028)

12. (fruit$ adj2 (intake or consum$ or increase or portion$ or serving$ or frequenc$ or number$ or preference$ or choice$)).ti,ab,kw. (10,978)

13. (vegetable$ adj2 (intake or consum$ or increase or portion$ or serving$ or frequenc$ or number$ or preference$ or choice$)).ti,ab,kw. (11,216)

14. “5 a day”.ti,ab,kw. (220)

15. “five a day”.ti,ab,kw. (61)

16. ((food or diet$) adj (choice$ or frequenc$ or intake)).ti,ab,kw. (96,918)

17. Feeding behavior/ (74,811)

18. Food preferences/ (10,664)

19. exp diet/ (282,043)

20. obesity/ (370,339)

21. 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 (728,864)

22. (editorial or letter).pt. (1,616,116)

23. 21 not 22 (698,752)

24. (physical adj3 (fit$ or train$ or activ$ or endur$)).ti,ab,kw. (153,242)

25. (exercis$ adj3 (fit$ or train$ or activ$ or endur$)).ti,ab,kw. (47,156)

26. ((promot$ or uptak$ or encourag$ or increas$ or start$ or adher$) adj3 (exercis$ or gym$ or sport$ or fitness)).ti,ab,kw. (39,955)

27. ((decreas$ or reduc$ or discourag$) adj3 (sedentary or deskbound)).ti,ab,kw. (1664)

28. (sport$ or walk$ or running or jogging or bicycling or biking or swimming).ti,ab,kw. (311,717)

29. (active adj (travel$ or transport$ or commut$)).ti,ab,kw. (9344)

30. exp fitness/ (34,746)

31. exp recreation/ (61,858)

32. running/ (24,432)

33. jogging/ (1834)

34. swimming/ (19,775)

35. exp walking/ (101,072)

36. 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 or 35 (597,469)

37. (editorial or letter).pt. (1,616,116)

38. 36 not 37 (585,776)

39. exp smoking/ (343,129)

40. smoking cessation/ (52,639)

41. smoking reduction/ (65)

42. (smoking or antismoking or anti-smoking).ti,ab,kw. (283,613)

43. (smoker or smokers).ti,ab,kw. (116,016)

44. nicotine/ (41,696)

45. tobacco/ (41,521)

46. “tobacco use”/ (7462)

47. tobacco smoke/ (12,136)

48. electronic cigarette/ (3860)

49. (electronic cigarette$ or e-cigarette$ or e-cig or e-cigs).ti,ab,kw. (3825)

50. (vape or vapes or vaper or vapers or vaping).ti,ab,kw. (625)

51. 39 or 40 or 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 (486,274)

52. (editorial or letter).pt. (1,616,116)

53. 51 not 52 (462,985)

54. drinking behavior/ (44,511)

55. alcohol intoxication/ (11,353)

56. exp alcoholic beverage/ (25,943)

57. drinking behavior/ (44,511)

58. (beer$ or wine$ or cider$ or alcopop$ or spirit or spirits).ti,ab,kw. (42,083)

59. (alcohol$ adj2 (binge or excessive$ or harm$ or heavily or heavy or misus$ or abus$ or consum$)).ti,ab,kw. (90,980)

60. (drink$ adj2 (binge or excessive$ or harm$ or heavily or heavy or misus$ or abus$ or consum$)).ti,ab,kw. (23,956)

61. (intoxicat$ or inebriat$ or drunk$).ti,ab,kw. (60,002)

62. 54 or 55 or 56 or 57 or 58 or 59 or 60 or 61 (229,090)

63. (editorial or letter).pt. (1,616,116)

64. 62 not 63 (223,747)

65. exp drug dependence/ (210,017)

66. inhalant abuse/ (504)

67. cannabis addiction/ (8769)

68. ((drug$ or substance$) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (264,725)

69. ((heroin or opiate$ or cocaine or crack) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (21,470)

70. ((cannabis or marijuana) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (15,565)

71. ((benzodiazepine$ or amphetamine$ or methamphetamine$ or MDMA or ecstasy) adj2 (use$ or using or abuse$ or abusing or misuse$ or misuising)).ti,ab,kw. (11,819)

72. (solvent$ adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (12,664)

73. (street drug$ or recreational drug$ or illicit drug$).ti,ab,kw. (17,283)

74. (prescri$ drug$ adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (3309)

75. polydrug use$.ti,ab,kw. (1032)

76. inject$ drug use$.ti,ab,kw. (12,111)

77. (needle adj3 shar$).ti,ab,kw. (1633)

78. (syringe$ adj3 shar$).ti,ab,kw. (877)

79. 65 or 66 or 67 or 68 or 69 or 70 or 71 or 72 or 73 or 74 or 75 or 76 or 77 or 78 (464,270)

80. (editorial or letter).pt. (1,616,116)

81. 79 not 80 (449,328)

82. 23 or 38 or 53 or 64 or 81 (2,132,733)

83. 10 and 82 (49,143)

84. clinical trial/ (950,305)

85. randomized controlled trial/ (517,116)

86. randomization/ (79,616)

87. single blind procedure/ (32,661)

88. double blind procedure/ (153,895)

89. crossover procedure/ (56,850)

90. placebo/ (324,536)

91. randomi?ed controlled trial$.tw. (187,656)

92. rct.tw. (29,684)

93. random allocation.tw. (1857)

94. randomly allocated.tw. (30,836)

95. allocated randomly.tw. (2373)

96. (allocated adj2 random).tw. (875)

97. single blind$.tw. (21,697)

98. double blind$.tw. (191,182)

99. ((treble or triple) adj blind$).tw. (846)

100. placebo$.tw. (278,797)

101. prospective study/ (475,629)

102. or/84-101 (1,953,083)

103. case study/ (56,829)

104. case report.tw. (364,419)

105. abstract report/or letter/ (1,074,387)

106. or/103-105 (1,486,683)

107. 102 not 106 (1,903,347)

108. 83 and 107 (6687)

109. (animal/or animal experiment/or animal model/or animal tissue/or nonhuman/) not exp human/(5,563,208)

110. (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys).ti. (1,984,529)

111. editorial.pt. or case report.ti. (838,717)

112. 109 or 110 or 111 (6,771,530)

113. 108 not 112 (6650)

114. (lifestyle adj2 (intervention$ or program$)).ti,ab,kw. (10,866)

115. (life style adj2 (intervention$ or program$)).ti,ab,kw. (523)

116. (behavior$ change adj (intervention$ or program$)).ti,ab,kw. (931)

117. (behaviour$ change adj (intervention$ or program$)).ti,ab,kw. (727)

118. (risk factor adj2 (program$ or intervention$)).ti,ab,kw. (1267)

119. (single risk factor adj2 (program$ or intervention$)).ti,ab,kw. (13)

120. (multifactorial lifestyle adj (intervention$ or program$)).ti,ab,kw. (22)

121. (health behavior$ adj (program$ or intervention$)).ti,ab,kw. (239)

122. (health behaviour$ adj (program$ or intervention$)).ti,ab,kw. (87)

123. (health risk behavior$ adj (program$ or intervention$)).ti,ab,kw. (5)

124. (health risk behaviour$ adj (program$ or intervention$)).ti,ab,kw. (0)

125. (risk behavior$ adj (program$ or intervention$)).ti,ab,kw. (70)

126. (risk behaviour$ adj (program$ or intervention$)).ti,ab,kw. (11)

127. health behavior$ change intervention$.ti,ab,kw. (108)

128. health behaviour$ change intervention$.ti,ab,kw. (52)

129. behavior$ risk factor$ intervention$.ti,ab,kw. (4)

130. behaviour$ risk factor$ intervention$.ti,ab,kw. (0)

131. behavior$ risk factor$ program$.ti,ab,kw. (0)

132. behaviour$ risk factor$ program$.ti,ab,kw. (0)

133. risk behaviour$ intervention$.ti,ab,kw. (9)

134. risk behavior$ intervention$.ti,ab,kw. (60)

135. risk behaviour$ program$.ti,ab,kw. (2)

136. risk behavior$ program$.ti,ab,kw. (10)

137. 114 or 115 or 116 or 117 or 118 or 119 or 120 or 121 or 122 or 123 or 124 or 125 or 126 or 127 or 128 or 129 or 130 or 131 or 132 or 133 or 134 or 135 or 136 (14,472)

138. 83 and 137 (222)

139. (animal/or animal experiment/or animal model/or animal tissue/or nonhuman/) not exp human/ (5,563,208)

140. (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys).ti. (1,984,529)

141. editorial.pt. or case report.ti. (838,717)

142. 139 or 140 or 141 (6,771,530)

143. 138 not 142 (222)

144. qualitative research/ (56,931)

145. qualitative.ti,ab. (231,093)

146. patient attitude/ (61,649)

147. feasibility study/ (98,850)

148. (accept$ or attitude$ or barrier$ or belief$ or believ$ or consider$ or experienc$ or facilitat$ or feasib$ or implement$ or option$ or preference$ or promot$ or view or views or viewpoint$).ti. (949,167)

149. 144 or 145 or 146 or 147 or 148 (1,269,494)

150. 83 and 149 (2676)

151. (animal/or animal experiment/or animal model/or animal tissue/or nonhuman/) not exp human/ (5,563,208)

152. (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys).ti. (1,984,529)

153. editorial.pt. or case report.ti. (838,717)

154. 151 or 152 or 153 (6,771,530)

155. 150 not 154 (2590)

156. 113 or 143 or 155 (8976)

157. limit 156 to yr=“1990 -Current” (8722)

158. limit 157 to embase (6199).

Search strategy

1. exp Schizophrenia/ (98,539)

2. Bipolar Disorder/ (37,622)

3. exp Psychotic Disorders/ (48,550)

4. exp Affective Disorders, Psychotic/ (2210)

5. ((acute or chronic$ or persistent$ or serious$ or severe$) adj2 (mental$ or psych$) adj2 (disease$ or disorder$ or disturbanc$ or ill$)).ti,ab,kw. (13,950)

6. (SMI adj3 (individual$ or inpatient$ or in-patient$ or outpatient$ or out-patient$ or patient$ or people or person$ or population$)).ti,ab,kw. (1040)

7. ((mental health or psychiatric) adj2 (inpatient$ or in-patient$)).ti,ab,kw. (9545)

8. schizophren$.ti,ab,kw. (117,306)

9. (psychotic or psychosis or psychoses).ti,ab,kw. (61,647)

10. bipolar.ti,ab,kw. (56,740)

11. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 or 10 (259,622)

12. (healthy adj2 (diet$ or eating)).ti,ab,kw. (10,696)

13. (fruit$ adj2 (intake or consum$ or increase or portion$ or serving$ or frequenc$ or number$ or preference$ or choice$)).ti,ab,kw. (8756)

14. (vegetable$ adj2 (intake or consum$ or increase or portion$ or serving$ or frequenc$ or number$ or preference$ or choice$)).ti,ab,kw. (8962)

15. “5 a day”.ti,ab,kw. (167)

16. “five a day”.ti,ab,kw. (44)

17. ((food or diet$) adj (choice$ or frequenc$ or intake)).ti,ab,kw. (73,165)

18. Feeding Behavior/ (75,218)

19. Food preferences/ (13,009)

20. nutrition therapy/or exp diet therapy/or exp diet/ (264,431)

21. obesity/or overweight/ (172,728)

22. 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 or 21 (506,530)

23. (comment or editorial or letter).pt. (1,663,957)

24. 22 not 23 (485,772)

25. (physical adj3 (fit$ or train$ or activ$ or endur$)).ti,ab,kw. (110,746)

26. (exercis$ adj3 (fit$ or train$ or activ$ or endur$)).ti,ab,kw. (34,892)

27. ((promot$ or uptak$ or encourag$ or increas$ or start$ or adher$) adj3 (exercis$ or gym$ or sport$ or fitness)).ti,ab,kw. (31,395)

28. ((decreas$ or reduc$ or discourag$) adj3 (sedentary or deskbound)).ti,ab,kw. (1308)

29. (sport$ or walk$ or running or jogging or bicycling or biking or swimming).ti,ab,kw. (239,832)

30. (active adj (travel$ or transport$ or commut$)).ti,ab,kw. (8634)

31. physical fitness/ (25,838)

32. exp Recreation/ (189,425)

33. exp Exercise Therapy/or exp exercise/ (199,415)

34. running/or jogging/or swimming/or walking/ (61,927)

35. 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 or 33 or 34 (540,109)

36. (letter or editorial or comment).pt. (1,663,957)

37. 35 not 36 (521,636)

38. exp smoking/ (137,824)

39. Smoking Cessation/or Smoking Reduction/ (26,021)

40. (smoking or antismoking or anti-smoking).ti,ab,kw. (201,198)

41. (smoker or smokers).ti,ab,kw. (78,963)

42. exp Nicotine/ (24,185)

43. exp Tobacco/ (28,846)

44. exp Tobacco Smoking/ (878)

45. “Tobacco Use Disorder”/ (10,429)

46. “Tobacco Use Cessation”/ (1034)

47. Tobacco, Smokeless/ (3414)

48. Tobacco Smoke Pollution/ (12,348)

49. Electronic Nicotine Delivery Systems/ (2007)

50. (electronic cigarette$ or e-cigarette$ or e-cig or e-cigs).ti,ab,kw. (3440)

51. (vape or vapes or vaper or vapers or vaping).ti,ab,kw. (594)

52. 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 or 46 or 47 or 48 or 49 or 50 or 51 (314,465)

53. (letter or editorial or comment).pt. (1,663,957)

54. 52 not 53 (301,222)

55. exp Alcohol Drinking/ (63,656)

56. exp Alcoholic Intoxication/ (12,115)

57. exp Alcoholic Beverages/ (18,386)

58. exp Drinking Behavior/ (69,908)

59. (beer$ or wine$ or cider$ or alcopop$ or spirit or spirits).ti,ab,kw. (33,610)

60. alcohol$.ti,ab,kw. (305,539)

61. (drink$ adj2 (binge or excessive$ or harm$ or heavily or heavy or misus$ or abus$ or consum$)).ti,ab,kw. (17,180)

62. (intoxicat$ or inebriat$ or drunk$).ti,ab,kw. (47,727)

63. 55 or 56 or 57 or 58 or 59 or 60 or 61 or 62 (395,310)

64. (comment or editorial or letter).pt. (1,663,957)

65. 63 not 64 (384,701)

66. substance-related disorders/or inhalant abuse/or marijuana abuse/or substance abuse, intravenous/ (107,638)

67. Drug Users/ (2555)

68. ((drug$ or substance$) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (191,716)

69. ((heroin or opiate$ or cocaine or crack) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (16,274)

70. ((cannabis or marijuana) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (11,836)

71. ((benzodiazepine$ or amphetamine$ or methamphetamine$ or MDMA or ecstasy) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (8864)

72. (solvent$ adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (9608)

73. (street drug$ or recreational drug$ or illicit drug$).ti,ab,kw. (12,452)

74. (prescri$ drug$ adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (2241)

75. polydrug use$.ti,ab,kw. (849)

76. inject$ drug use$.ti,ab,kw. (9504)

77. (needle adj3 shar$).ti,ab,kw. (1311)

78. (syringe$ adj3 shar$).ti,ab,kw. (726)

79. or/66-78 (272,714)

80. (letter or editorial or comment).pt. (1,663,957)

81. 79 not 80 (262,481)

82. 24 or 37 or 54 or 65 or 81 (1,748,549)

83. 11 and 82 (27,989)

84. randomized controlled trial.pt. (469,898)

85. controlled clinical trial.pt. (92,702)

86. randomized.ab. (424,124)

87. placebo.ab. (192,498)

88. clinical trials as topic.sh. (185,008)

89. randomly.ab. (298,839)

90. trial.ti. (188,739)

91. 84 or 85 or 86 or 87 or 88 or 89 or 90 (1,177,116)

92. exp animals/not humans.sh. (4,506,067)

93. 91 not 92 (1,083,038)

94. 83 and 93 (2417)

95. (lifestyle adj2 (intervention$ or program$)).ti,ab,kw. (7301)

96. (life style adj2 (intervention$ or program$)).ti,ab,kw. (266)

97. (behavior$ change adj (intervention$ or program$)).ti,ab,kw. (897)

98. (behaviour$ change adj (intervention$ or program$)).ti,ab,kw. (629)

99. (risk factor adj2 (program$ or intervention$)).ti,ab,kw. (1029)

100. (single risk factor adj2 (program$ or intervention$)).ti,ab,kw. (12)

101. (multifactorial lifestyle adj (intervention$ or program$)).ti,ab,kw. (14)

102. (health behavior$ adj (program$ or intervention$)).ti,ab,kw. (278)

103. (health behaviour$ adj (program$ or intervention$)).ti,ab,kw. (87)

104. (health risk behavior$ adj (program$ or intervention$)).ti,ab,kw. (6)

105. (health risk behaviour$ adj (program$ or intervention$)).ti,ab,kw. (0)

106. (risk behavior$ adj (program$ or intervention$)).ti,ab,kw. (98)

107. (risk behaviour$ adj (program$ or intervention$)).ti,ab,kw. (15)

108. health behavior$ change intervention$.ti,ab,kw. (105)

109. health behaviour$ change intervention$.ti,ab,kw. (45)

110. behavior$ risk factor$ intervention$.ti,ab,kw. (2)

111. behaviour$ risk factor$ intervention$.ti,ab,kw. (0)

112. behavior$ risk factor$ program$.ti,ab,kw. (0)

113. behaviour$ risk factor$ program$.ti,ab,kw. (0)

114. risk behaviour$ intervention$.ti,ab,kw. (11)

115. risk behavior$ intervention$.ti,ab,kw. (47)

116. risk behaviour$ program$.ti,ab,kw. (2)

117. risk behavior$ program$.ti,ab,kw. (9)

118. 95 or 96 or 97 or 98 or 99 or 100 or 101 or 102 or 103 or 104 or 105 or 106 or 107 or 108 or 109 or 110 or 111 or 112 or 113 or 114 or 115 or 116 or 117 (10,433)

119. 11 and 82 and 118 (137)

120. 94 or 119 (2513)

121. limit 120 to yr=“1990 -Current” (2312)

122. qualitative research/ (41,870)

123. qualitative.ti,ab. (190,253)

124. “Patient Acceptance of Health Care”/ (40,335)

125. Feasibility Studies/ (59,810)

126. (accept$ or attitude$ or barrier$ or belief$ or believ$ or consider$ or experienc$ or facilitat$ or feasib$ or implement$ or option$ or preference$ or promot$ or view or views or viewpoint$).ti. (817,156)

127. 122 or 123 or 124 or 125 or 126 (1,056,514)

128. 83 and 127 (1538)

129. exp animals/not humans.sh. (4,506,067)

130. 128 not 129 (1491)

131. limit 130 to yr=“1990 -Current” (1331)

132. 121 or 131 (3450).

Search strategy

1. exp Schizophrenia/ (85,600)

2. Bipolar Disorder/ (25,219)

3. exp Psychosis/ (109,658)

4. ((acute or chronic$ or persistent$ or serious$ or severe$) adj2 (mental$ or psych$) adj2 (disease$ or disorder$ or disturbanc$ or ill$)).ti,ab,kw. (16,838)

5. (SMI adj3 (individual$ or inpatient$ or in-patient$ or outpatient$ or out-patient$ or patient$ or people or person$ or population$)).ti,ab,kw. (931)

6. ((mental health or psychiatric) adj2 (inpatient$ or in-patient$)).ti,ab,kw. (11,386)

7. schizophren$.ti,ab,kw. (113,900)

8. (psychotic or psychosis or psychoses).ti,ab,kw. (68,011)

9. bipolar.ti,ab,kw. (36,173)

10. 1 or 2 or 3 or 4 or 5 or 6 or 7 or 8 or 9 (209,491)

11. (healthy adj2 (diet$ or eating)).ti,ab,kw. (3697)

12. (fruit$ adj2 (intake or consum$ or increase or portion$ or serving$ or frequenc$ or number$ or preference$ or choice$)).ti,ab,kw. (1871)

13. (vegetable$ adj2 (intake or consum$ or increase or portion$ or serving$ or frequenc$ or number$ or preference$ or choice$)).ti,ab,kw. (2298)

14. “5 a day”.ti,ab,kw. (82)

15. “five a day”.ti,ab,kw. (12)

16. ((food or diet$) adj (choice$ or frequenc$ or intake)).ti,ab,kw. (14,493)

17. Eating Attitudes/ (1530)

18. Food preferences/ (4452)

19. Diets/or Food Intake/ (24,024)

20. obesity/or overweight/ (23,672)

21. 11 or 12 or 13 or 14 or 15 or 16 or 17 or 18 or 19 or 20 (56,824)

22. 10 and 21 (1093)

23. (physical adj3 (fit$ or train$ or activ$ or endur$)).ti,ab,kw. (34,827)

24. (exercis$ adj3 (fit$ or train$ or activ$ or endur$)).ti,ab,kw. (5926)

25. ((promot$ or uptak$ or encourag$ or increas$ or start$ or adher$) adj3 (exercis$ or gym$ or sport$ or fitness)).ti,ab,kw. (5944)

26. ((decreas$ or reduc$ or discourag$) adj3 (sedentary or deskbound)).ti,ab,kw. (438)

27. (sport$ or walk$ or running or jogging or bicycling or biking or swimming).ti,ab,kw. (72,925)

28. (active adj (travel$ or transport$ or commut$)).ti,ab,kw. (646)

29. physical fitness/ (4026)

30. exp Recreation/ (33,080)

31. exp Exercise Therapy/or exp exercise/ (24,104)

32. Physical Activity/ (17,162)

33. 23 or 24 or 25 or 26 or 27 or 28 or 29 or 30 or 31 or 32 (144,672)

34. 10 and 33 (2237)

35. exp Tobacco Smoking/ (29,372)

36. Smoking Cessation/ (12,191)

37. (smoking or antismoking or anti-smoking).ti,ab,kw. (44,133)

38. (smoker or smokers).ti,ab,kw. (20,066)

39. Nicotine/ (10,413)

40. Smokeless Tobacco/ (773)

41. Passive Smoking/ (776)

42. Electronic Cigarettes/ (730)

43. Tobacco, Smokeless/ (773)

44. (electronic cigarette$ or e-cigarette$ or e-cig or e-cigs).ti,ab,kw. (1138)

45. (vape or vapes or vaper or vapers or vaping).ti,ab,kw. (175)

46. 35 or 36 or 37 or 38 or 39 or 40 or 41 or 42 or 43 or 44 or 45 (55,798)

47. 10 and 46 (2676)

48. exp Alcohol Abuse/ (46,611)

49. exp Alcohol Drinking Patterns/ (63,984)

50. exp Alcohol Intoxication/ (3029)

51. (beer$ or wine$ or cider$ or alcopop$ or spirit or spirits).ti,ab,kw. (16,133)

52. (alcohol$ adj2 (binge or excessive$ or harm$ or heavily or heavy or misus$ or abus$ or consum$)).ti,ab,kw. (34,787)

53. (drink$ adj2 (binge or excessive$ or harm$ or heavily or heavy or misus$ or abus$ or consum$)).ti,ab,kw. (12,209)

54. (intoxicat$ or inebriat$ or drunk$).ti,ab,kw. (11,855)

55. 48 or 49 or 50 or 51 or 52 or 53 or 54 (99,930)

56. 10 and 55 (5834)

57. exp Drug Abuse/ (105,866)

58. exp Drug Addiction/ (13,397)

59. Heroin Addiction/ (2508)

60. ((drug$ or substance$) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (108,973)

61. ((heroin or opiate$ or cocaine or crack) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (10,568)

62. ((cannabis or marijuana) adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (10,161)

63. ((benzodiazepine$ or amphetamine$ or methamphetamine$ or MDMA or ecstasy) adj2 (use$ or using or abuse$ or abusing or misuse$ or misuising)).ti,ab,kw. (5549)

64. (solvent$ adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (270)

65. (street drug$ or recreational drug$ or illicit drug$).ti,ab,kw. (7932)

66. (prescri$ drug$ adj2 (use$ or using or abuse$ or abusing or misuse$ or misusing)).ti,ab,kw. (1169)

67. polydrug use$.ti,ab,kw. (834)

68. inject$ drug use$.ti,ab,kw. (4229)

69. (needle adj3 shar$).ti,ab,kw. (584)

70. (syringe$ adj3 shar$).ti,ab,kw. (387)

71. 57 or 58 or 59 or 60 or 61 or 62 or 63 or 64 or 65 or 66 or 67 or 68 or 69 or 70 (175,902)

72. 10 and 71 (16,517)

73. 22 or 34 or 47 or 56 or 72 (22,832)

74. clinical trials/ (11,117)

75. clinical trial.md. (21,833)

76. placebo/ (5150)

77. control$.ti,ab. (635,515)

78. random$.ti,ab. (181,628)

79. exp treatment/ (720,498)

80. 74 or 75 or 76 or 77 or 78 or 79 (1,330,498)

81. 73 and 80 (11,362)

82. (lifestyle adj2 (intervention$ or program$)).ti,ab,kw. (1742)

83. (life style adj2 (intervention$ or program$)).ti,ab,kw. (40)

84. (behavior$ change adj (intervention$ or program$)).ti,ab,kw. (702)

85. (behaviour$ change adj (intervention$ or program$)).ti,ab,kw. (217)

86. (risk factor adj2 (program$ or intervention$)).ti,ab,kw. (126)

87. (single risk factor adj2 (program$ or intervention$)).ti,ab,kw. (3)

88. (multifactorial lifestyle adj (intervention$ or program$)).ti,ab,kw. (3)

89. (health behavior$ adj (program$ or intervention$)).ti,ab,kw. (182)

90. (health behaviour$ adj (program$ or intervention$)).ti,ab,kw. (26)

91. (health risk behavior$ adj (program$ or intervention$)).ti,ab,kw. (2)

92. (health risk behaviour$ adj (program$ or intervention$)).ti,ab,kw. (0)

93. (risk behavior$ adj (program$ or intervention$)).ti,ab,kw. (47)

94. (risk behaviour$ adj (program$ or intervention$)).ti,ab,kw. (4)

95. health behavior$ change intervention$.ti,ab,kw. (101)

96. health behaviour$ change intervention$.ti,ab,kw. (25)

97. behavior$ risk factor$ intervention$.ti,ab,kw. (2)

98. behaviour$ risk factor$ intervention$.ti,ab,kw. (0)

99. behavior$ risk factor$ program$.ti,ab,kw. (0)

100. behaviour$ risk factor$ program$.ti,ab,kw. (0)

101. risk behaviour$ intervention$.ti,ab,kw. (3)

102. risk behavior$ intervention$.ti,ab,kw. (43)

103. risk behaviour$ program$.ti,ab,kw. (1)

104. risk behavior$ program$.ti,ab,kw. (4)

105. 82 or 83 or 84 or 85 or 86 or 87 or 88 or 89 or 90 or 91 or 92 or 93 or 94 or 95 or 96 or 97 or 98 or 99 or 100 or 101 or 102 or 103 or 104 (3008)

106. 10 and 105 (179)

107. 81 or 106 (11,482)

108. qualitative research/ (7974)

109. qualitative.ti,ab. (143,449)

110. (accept$ or attitude$ or barrier$ or belief$ or believ$ or consider$ or experienc$ or facilitat$ or feasib$ or implement$ or option$ or preference$ or promot$ or view or views or viewpoint$).ti. (298,416)

111. 108 or 109 or 110 (416,644)

112. 73 and 111 (1132)

113. 107 or 112 (12,057)

114. animal models/ (30,681)

115. (animal not (animal and human)).po. (346,436)

116. (rat or rats or mouse or mice or swine or porcine or murine or sheep or lambs or pigs or piglets or rabbit or rabbits or cat or cats or dog or dogs or cattle or bovine or monkey or monkeys).ti. (143,111)

117. 114 or 115 or 116 (366,257)

118. 113 not 117 (11,736)

119. limit 118 to yr=“1990 -Current” (10,354).

Search strategies

Serious mental illness and risk behaviours and poor diet

Records identified: 48.

Serious mental illness and risk behaviours and lack of physical activity

Records identified: 127.

Serious mental illness and risk behaviours and smoking

Records identified: 68.

Serious mental illness and risk behaviours and alcohol use

Records identified: 141.

Serious mental illness and risk behaviours and drug misuse

Records identified: 257.

Serious mental illness and lifestyle interventions

Records identified: 17.