Author, date
Type of review (K)
Sample description
Intervention variables
Outcome measures
Types of research design
Weight management findings
Comments
Bonfioli et al., 2012 [20]
SR and MA (13)
Adults 18–64, at least 50 % with SZ-like illness, bipolar disorder, or depression with psychotic features
Psychoeducational , cognitive-behavioural, nutritional or physical activity-based interventions, aimed at weight reduction or prevention of weight gain
BMI
RCT
Compared to controls, experimental groups showed a mean BMI reduction of −0.98 kg/m2 (95 % CI: −1.31 to −0.65 kg/m2). Equivalent to loss of 3.12 % of initial weight. Prevention studies with individual psychoeducational programmes that include diet and/or physical activity have the highest impact
Cimo et al., 2012 [21]
SR (4) (One of these was a follow-up to previous study)
Type 2 diabetes and SZ/SZA
Must target a lifestyle factor associated with diabetes self-care, such as problem-solving skills, education classes, diet, or exercise
HbA1c, fasting blood glucose (FBG) , BMI or weight lost (measured in pounds or kilograms)
Various
Only one study reported weight loss in an RCT
Low number of included studies, and inclusion of a variety of study designs makes it difficult to draw conclusions
Authors conclude diabetes education is effective when it incorporates diet and exercise components. Weight loss is possible
Firth et al., 2015 [22]
SR (20) and MA (11)
Non-affective psychotic disorders (but including SZA) or first episode psychosis
Exercise
At least one quantitative measure of physical or mental health
Interventions (RCT, non-randomized controlled trial, case series/uncontrolled longitudinal)
Body weight/BMI studies in SR (K = 9) and MA (K = 4). Exercise did not significantly reduce BMI in meta-analysis: mean difference = −0.98 kg/m2; 95 % CI −3.17 to 1.22 kg/m2. Systematic review results of other studies were mixed
Only RCT for MA
Gierisch et al., 2014 [23]
SR (33) and MA (11) (10 Behavioural)
Adults with SMI
Patient-focused behavioral interventions/peer or family support intervention/pharmacological treatments targeting weight control/glucose levels/lipid levels/overall CVD risk
Key outcomes were weight (kg), glycosylated haemoglobin A1c (HbA1c), total and low-density lipoprotein (LDL) cholesterol
RCT
Behavioural significantly reduced BMI in meta-analysis of behavioural interventions (K = 10): mean difference = −3.14 kg/m2; 95 % CI −4.33 to −1.96
Metformin (K = 5, mean difference = −4.13 kg/m2; 95 % CI −6.58 to −1.68) and anti-convulsant medication (K = 4, mean difference = −5.11 kg/m2; 95 % CI −9.48 to −0.74) also effective pharmacological interventions
No peer or family support studies; Insufficient strength of evidence for any intervention having effect on glucose control (glycosylated haemoglobin) or lipid control
Hjorth et al., 2014 [24]
SR (23)
>50 % SZ, SZA or schizophreniform disorder
Non-pharmacological interventions (diet, exercise, cognitive behavioural therapy, or combined) aimed at weight reduction/reducing physical illness compared to with standard care
Weight loss/weight maintenance, other physical health outcomes
Interventions (Randomized and Non-randomized clinically controlled)
Diet (K = 4) three studies demonstrated improved weight management; one study offering free fruits and vegetables did not show any health parameter change
Two diet interventions included walking, but were not considered “combinations”
Exercise (K = 5): four studies reported weight loss/better weight management with exercise, one did not report weight outcomes but reported improved cardiovascular fitness
One pedometer and motivational interviewing, one fitness training and nutritional advice
Cognitive/Behavioural- Therapy (K = 3): two studies showed significant better weight management; one not significant at study end, but significant at 6-month follow-up
Combined (K = 11): improved weight control in ten studies (two of these reported improved metabolic profile), remaining study demonstrated reduction in metabolic syndrome (weight not reported in review)
Pearsall et al., 2016 [25]
SR and MA (0)
Any age, schizophrenia, or other types of schizophrenia-like psychosis
dietary advice, with the aim of changing and improving dietary intake; only diet, no exercise/supplements
Nutritional intake; BMI/weight
RCT
No studies eligible for inclusion
Cochrane Review
Pearsall et al., 2014 [26]
SR and MA (8)—Body weight/BMI in SR (4) and MA (4)
Adults with schizophrenia, schizoaffective disorders, or bipolar-affective disorder
Promote exercise or physical activity
Any outcome (not specified in methods, post hoc analysis)
RCT
MA showed no difference in BMI (K = 4) or weight (K = 2) between standard care and exercise at end of trials. Change in BMI/weight not in MA, but SR reported one study showed reduction in body weight greater in intervention than control, other study no significant difference. Change in BMI for only one study reported in SR, with no significant difference between intervention and control
MA appears to just evaluate final BMI and weight, not change. This does not account for initial differences between groups
Rosen-baum et al., 2014 [27]
SR and MA (39) (12 schizophrenia, 1 bipolar disorder, 1 first episode psychosis; 1 severe mental illness); (11 for anthropometric measures)
18 years of age or older, in whom a DSM or ICD diagnosis of mental illness was made. Dysthymia, “mild-depression”, and eating disorders were excluded
Any form of physical activity
Various health outcomes, including BMI/Weight
RCT
Pooled anthropometric measures indicate effect in favour of exercise: SMD = 0.24; 95 % CI, 0.06–0.41; p < 0.05; I 2 = 0 %
8 of 11 studies analysed in anthropometric MA described sample as SMI
Whitney et al., 2015 [28]
SR (17)
>50 % of participants treated with clozapine (all treated with antipsychotics)
Pharmacological and non-pharmacological weight management interventions
BMI/Weight
Various
Non-Pharmacological (n = 2): Only one intervention study which found significant reduction in weight (−5.2 kg) and BMI with diet and exercise intervention versus control
Combination (K = 1): Sibutramine + behavioural and nutritional counselling found no change in weight
Caem-merer et al., 2012 [29]
SR and MA (17)
Implied: individuals with antipsychotic weight gain
Any non-pharmacological interventions to address antipsychotic associated weight
BMI /Weight
RCT
Interventions led to a significant reduction in weight (−3.12 kg; CI:−4.03, −2.21, p < 0.001) and BMI (−0.94 kg/m2; CI:−1.45, −0.43, p < 0.001) compared with control groups
Prevention and intervention trial sub-analyses performed, similar to pooled results
Results were mixed among the seven reviews that performed meta-analyses. Two found no significant difference between interventions and controls [22, 26], four found an effect in favour of interventions [20, 23, 27, 29], while one did not find any studies that met the inclusion criteria [25]. Both meta-analyses that found no effect for the intervention included interventions focused on physical activity. The only other meta-analysis to limit interventions solely to physical activity did find a positive effect in favour of physical activity, but included studies that were not limited to SMI [27]. These results indicate that exercise only interventions are unlikely to assist weight management. Rather multifaceted and comprehensive weight management programmes that aim to improve diet, increase physical activity, and provide counselling are likely to be more successful in improving weight management.
The reviews conducted by Bonfioli et al. [20], Caemmerer et al. [29], and Gierisch et al. [23] could be considered the most comprehensive due to their inclusion of any non-pharmacological intervention to target weight management. These reviews are also specific to people with SMI , rather than all mental illnesses , and each provide meta-analyses of randomized controlled trials providing high quality level of evidence. Caemmerer et al.’s [29] review meta-analysed 17 studies, the majority of which (nine) were described as cognitive behavioural therapy or a psychoeducational programme , and another four were described as a combination of nutrition and exercise. Treatment programmes lasted 8–72 weeks. The average treatment period was 19.6 weeks. Overall, the weighted mean difference across all trials was significant for weight (−3.12 kg, CI: −4.03 to −2.21, p < 0.001) and BMI (−0.94 kg/m2, CI: −1.45 to −0.43, p < 0.001), indicating that small reductions in weight are feasible with non-pharmacological therapies and these changes can be accrued within an approximate 6-month treatment period. Furthermore, this review provides evidence that weight loss can be maintained. Five programmes had follow-up periods 8–52 weeks (mean 14.4 weeks) after the treatment ended. Significantly greater weight loss persisted in favour of the intervention group compared to control (−3.48 kg, CI: −6.37 to −0.58, p = 0.02).
Caemmerer and colleagues [29] also performed several sub-analyses to identify moderating variables. Intervention trials to reduce weight gain among individuals taking antipsychotics were separated from prevention trials aiming to minimize weight gain after initiating antipsychotic medication. There was no significantly different effect for prevention (−3.23 kg, CI: −4.37 to −2.04, p < 0.001) versus treatment interventions (−3.12 kg, CI: −4.39 to −2.21, p < 0.001). Additional sensitivity analyses found comparable results between interventions >3 months and <3 months, as well as group-based and individual treatments. Cognitive-behavioural interventions had a smaller effect size than nutrition and/or exercise programmes though both were significantly better than controls. Finally, weight and BMI were significantly improved only in outpatient trials but not in inpatient or mixed samples. Overall, these analyses indicate that both prevention and intervention trials are roughly equivalent in effect, weight loss can be achieved in under 3 months, and that both group-based and individual interventions are similarly effective.
Gierisch and colleagues’ [23] review included ten behavioural interventions in their meta-analysis, all of which included either cognitive behavioural therapy or multiple lifestyle change factors. Pooled effects favoured behavioural interventions with a mean difference of −3.13 kg (CI: −4.21 to −2.05, p < 0.05). This review also reported positive effects for metformin (mean difference = −4.13 kg CI: −6.58 to −1.68, p < 0.05), and anticonvulsive medications topiramate and zonisamide (mean difference = −5.11 kg, CI: −9.48 to −0.74, p < 0.05). Bonfioli et al. [20] identified 17 studies in their systematic review, of which four were excluded from meta-analysis due to missing information. The remaining 13 studies all included elements of cognitive behavioural therapy or psychoeducation as opposed to diet or exercise alone. Overall, the findings were very similar to Caemmerer et al.’s [29] review, which found a weighted mean difference in BMI of −0.98 kg/m2 (CI: −1.31 to −0.65 kg/m2, p < 0.05), compared to control. In contrast to the findings reported by Caemmerer and colleagues [29], the sub-analyses reported by Bonfioli et al. [20] found that weight gain prevention studies were slightly more effective than treatment interventions, and group interventions were less effective than individual interventions. These differences in the results of sub-analyses reflect different inclusion criteria of the two reviews.
Taken together these three meta-analyses provide consistent evidence that lifestyle interventions that include diet and physical activity are feasible and are effective in reducing weight by approximately 3 kg or a BMI reduction of 0.9–1 kg/m2. Weight loss can also be maintained. However, the optimal duration and intensity (e.g. number of counselling sessions) of intervention is still unclear. At the least, meta-analyses suggest that optimal interventions should consider both diet and physical activity and include psychological/behavioural components. A more focused discussion is now provided concerning dietary and physical activity strategies that could be applied in the case of Amanda .
19.4 Intervening with Amanda
19.4.1 Nutrition
Nutrition interventions, as part of programmes that include regular physical activity and quality sleep, are an important component of healthy weight management. The dietary patterns of people with schizophrenia or those at high risk for psychosis are typically high in energy mainly contributed from excess fat (particularly saturated fat) and low in fruits, vegetables, and dietary fibre [30–34]. These eating patterns contribute both to dietary energy excess and gut microbiome alterations [35], which lead to subsequent weight gain, risks of conditions such as heart disease, type 2 diabetes, osteoarthritis, and exacerbation of mental health symptoms.
The causes of overeating are multifactorial [36]. Genetics, which can affect metabolic rate and weight status, and physiologic factors (e.g. leptin, ghrelen, uncoupling proteins, beta-endorphins, neuropeptide Y, decreased blood glucose) can reduce satiety or increase hunger. Recurrent and frequent binge eating can cause significant weight gain, and food environments, which offer an abundance of good-tasting, cheap food, makes it difficult for susceptible individuals to avoid such overeating triggers. Night eating, where most food energy is consumed between 8 p.m. and 6 a.m., is another common contributor to overeating.
For those with schizophrenia, other condition-related determinants contribute to excess food intake. As in Amanda’s case, psychiatric medications such as the second-generation antipsychotics, are associated with metabolic side effects of weight gain, dyslipidemia, insulin resistance, susceptibility to type 2 diabetes, and metabolic syndrome ([37–40]), which may be coupled by an a priori increased risk of obesity that has been reported in schizophrenia [41, 42]. Structural factors such as the built environment that controls food availability and social barriers such as negative discrimination and stigma can also contribute to the etiology of overeating and excess weight. Individual factors such as a history of trauma, memory or cognitive impairment, food-related hallucinations or delusions, comorbid eating disturbances, poor condition self-management skills, proneness to food fads and use of natural health products, concurrent substance use, social avoidance, symptoms of depression, anxiety, and poor sleep, and determinants of health such as low socioeconomic status, and food insecurity can create barriers to health services access and behavioural interventions [43–48].
Figure 19.1 outlines an algorithm that may be used in the assessment and management of weight and nutrition for individuals such as Amanda. Her initial assessment would include examining health status, nutrition-related biochemical measures and genomic markers where feasible, medication and natural health product use, food intake and eating behaviours, anthropometric measurements (e.g. height, weight, waist and hip circumferences), and health determinants such as income, food security, and living situation. Subsequently, an individualized lifestyle intervention would be negotiated where the nutrition component would focus on reducing energy intake in the range of 500–600 cal/day and applying cognitive and behavioural techniques aimed at dietary restructuring. The recommended macronutrient composition would emphasize high complex carbohydrates with moderate fat and protein levels where total energy is distributed as 20–30 % fat, 55–60 % carbohydrates, and 15–20 % protein to help reduce weight and control LDL-cholesterol, blood triglycerides, and blood glucose levels [49, 50], even in the context of antipsychotic medication use [45, 51]. The minimum caloric level for women such as Amanda is 1000 cal/day; for men no less than 1200 kcal/day should be consumed to ensure all nutrient requirements are being met. If the individual’s food intake is not balanced and varied, then fibre and micronutrients such as zinc, calcium, iron, and vitamin B12 may be lacking and supplementation may be required. Furthermore, should the person be pregnant or lactating, or have significant eating disturbances these lower calorie diets should be avoided. In designing a weight management plan for Amanda , a personalized non-complex approach is needed that incorporates three core strategies:
Fig. 19.1
Assessment and Weight Management Algorithm
- 1.
Determining readiness for change and setting realistic goals: Discussions about weight should be non-judgemental and explore the individual’s readiness for change [52]. If the individual has concerns about their weight, then goals should be set that are specific, reasonable, measurable, and promote gradual weight loss of 0.5–1 kg/week.
- 2.
Consuming less food energy: Nutrition education would focus on learning appropriate serving sizes, facilitating lowered energy intake by consuming less added and hidden fats, switching to low- and non-caloric beverages, and emphasizing consumption of vegetables and fruits to provide important micronutrients and promote satiety.
- 3.
Incorporate appropriate eating behaviour modifications: Exemplars of these include eating at set regular times in one location to avoid mindless eating, keeping a log of what is eaten, when, and why to identify social or emotional cues of overeating, saving high-calorie snacks foods for occasional special treats, avoiding purchases of problem foods, serving food portions on smaller dishes, and chewing foods slowly and stopping when feeling full.
As is the case with Amanda , several condition specific factors should be considered in the management of weight. While for some the structure of a diet plan may be helpful, therapeutic interventions may be needed alone or as adjunct to a diet plan to help correct eating disturbances and manage weight. Although a number of strategies may be used, evidence is currently lacking about their utility in mental health populations. For those who do not connect with hunger or satiety cues, mindful and intuitive eating counselling can help promote awareness of bodily sensations, self-care, and valuing health more than appearance [53, 54]. Motivational interviewing can also be used as a tool in weight management to help the individual resolve ambivalence about eating behaviour change and build intrinsic motivation [55]. Cognitive behaviour therapy (CBT) is a core strategy that can be used to address binge eating and improve body image [56]. Dialectical behaviour therapy is a form of CBT often used in more complex cases that can facilitate learning and skills development such as emotion regulation and distress intolerance that improve abilities to manage negative affect adaptively [57]. Should cognitive-behavioural and dialectical behaviour therapies not be effective, an alternative is acceptance and commitment therapy that can help the person work toward valued goals and life directions [58]. For individuals with cognitive impairments, cognitive adaptive training (CAT) may be used to help overcome challenges such as initiating and sustaining task behaviour (e.g. preparing meals). Exemplars of CAT environmental supports include separating foods into bins labelled by day of the week, using checklists, lists outlining steps for food preparation, and electronic auditory prompts, as well as practicing shopping at grocery stores using a food list [59]. While mobile electronic devices are popular tools to facilitate and maintain weight loss among overweight and obese populations, evidence on their sustained benefit and utility in mental health populations is currently lacking [60].
19.4.2 Physical Activity
Physical inactivity is itself a major cause of morbidity and mortality, and merits the same level of concern as other cardiovascular disease risk factors like smoking (e.g. [61]). Studies consistently highlight that individuals with schizophrenia are less active than the general population [62, 63]. This is reflected in the demonstration of significantly lower cardiorespiratory fitness among individuals with schizophrenia in comparison to the general population (e.g. [64]). Given the compelling evidence that regular physical activity is an effective preventative strategy against premature mortality, cardiovascular disease, stroke, hypertension, colon cancer, breast cancer, and type 2 diabetes for the general population [65], promoting physical activity is important irrespective of weight loss.
We have adapted a form of physical activity counselling that is designed to increase physical activity by addressing an individual’s needs, motivation, and barriers to activity [66]. Physical activity counselling draws from numerous theories of behaviour change including social cognitive theory and the transtheoretical model (TTM) [67, 68]. Social cognitive theory states that behaviour is influenced and influences both environmental and personal factors. Environmental factors include group structures, equipment, or various facilities. Personal factors include cognitions such as self-efficacy, mood, and attitudes. Together, these three determinants influence one another and form what is known as triadic reciprocal causation [67]. The TTM allows a clinician to assess whether an individual is ready to change a particular behaviour. Depending on their readiness to change, clients are placed into one of six stages: precontemplation (not considering changing behaviour), contemplation (intending to change behaviour in the next 6 months), preparation (intending to change behaviour in next 30 days), action (actively engaged in behaviour change for 6 months), maintenance (actively engaged in behaviour change for more than 6 months), and termination (actively engaged in behaviour change for more than 5 years). This then directs attention to those individuals who are contemplating or preparing to increase their physical activity. The TTM also provides stage specific strategies that strengthen a client’s self-efficacy and highlight the positive, rather than the negative, attributes of increasing physical activity. Self-efficacy has been shown to be an important mediator of physical activity behaviour change in people with schizophrenia [69].
The intervention was conducted over 2 months and included four 60 min weekly individual sessions that could be replicated:
Session one: Build rapport and gather knowledge. This first session provided an opportunity to assess current levels of physical activity and explore commitment to increasing physical activity. As a starting point, physical activity was assessed using the International Physical Activity Questionnaire (IPAQ; [70]). The IPAQ has previously been validated as a measure of physical activity behaviour for adults with schizophrenia [71]. Throughout the first session, various techniques were used to assess readiness to change. First, past, present, and future interests and physical activity were explored. Given Amanda’s prior interest in sport and exercise there is an excellent basis for establishing rapport. Additionally, confidence and importance scales were used to assess how confident each participant was about becoming more physically active and how important physical activity was to him or her [72, 73]. A similar approach could be taken in assessing interest in modifying dietary behaviour and a decision taken to focus on one or both behaviours. The first session provided a better understanding of why physical activity may or may not be important to the participant and what potential barriers existed. In addition to the confidence and importance scales, the participants completed a decisional balance exercise [72, 73]. This exercise helped participants explore and overcome their ambivalence to behaviour change by attempting to establish a discrepancy between the advantages and disadvantages of changing their behaviour.