An Australian experience – the Bondi keeping the body in mind (KBIM) program

The KBIM Program, developed at the Bondi Centre in South Eastern Sydney Local Health District, Australia, provides a pragmatic example of how lifestyle interventions can be successfully integrated within a community mental health service setting. KBIM evolved over a number of years, beginning with the implementation of routine metabolic screening in a first-episode psychosis (FEP) service (Curtis et al. Reference Curtis, Henry, Watkins, Newall, Samaras and Ward2011, Reference Curtis, Newall and Samaras2012). Mental health clinicians were seeing first-hand the rapid onset of substantial weight gain in large numbers of young people recently commenced on antipsychotic medication, which prompted a new focus on documenting the time-course of changes in key metabolic measures.

Following implementation of the screening program, clinicians began to explore novel strategies to counter the rapid weight gain frequently observed. Individual dietetic consultations were trialled, and found to be acceptable to the young FEP clients. An in-house gym was developed in the community centre, with an exercise physiologist (EP) working one-on-one with clients to prescribe and supervise exercise programs reflecting the fitness levels and goals of each young person. The multidisciplinary KBIM team included a senior nurse with experience in FEP and metabolic expertise, a dietician, an EP and a peer-support worker, with medical input from psychiatrists and an endocrinologist. The program included individual sessions with members of the KBIM team, along with the opportunity to participate in weekly sports groups, shopping tours and cooking groups with a focus on skill acquisition, in addition to access to an on-site gym supervised by both EPs and EP students.

With this combination of individualised and group-based interventions, antipsychotic-induced weight gain was prevented among young people experiencing FEP, when targeted interventions were delivered soon after antipsychotic initiation and as a core part of routine clinical services. Metabolic data were collected from a control group receiving care from an early psychosis program without access to the KBIM lifestyle interventions. The KBIM intervention group experienced significantly less weight gain compared with the standard care group (Curtis et al. Reference Curtis, Watkins, Rosenbaum, Teasdale, Kalucy, Samaras and Wardin press).

Following this successful pilot program, the local mental health district have funded the establishment of additional KBIM teams to provide lifestyle interventions to the entire mental health service in a staged rollout, first targeting early psychosis clients, those receiving clozapine and eventually all clients at an increased risk of cardio-metabolic disease. The KBIM intervention drew on already established evidence-based intervention strategies, for example individualised exercise and dietetic consultations (Rosenbaum et al. Reference Rosenbaum, Tiedemann, Sherrington, Curtis and Ward2014; Teasdale et al. Reference Teasdale, Harris, Rosenbaum, Watkins, Samaras, Curtis and Ward2015), which were integrated as routine elements of the mental health service. The KBIM experience provides an example of research translation, and the importance of obtaining ‘real-world’ evidence for treatment effectiveness that does not rely on RCT methodology.

An epidemiological perspective

In his seminal paper published over 30 years ago, Professor Geoffrey Rose proposed the concept of the prevention paradox; ‘a measure that brings large benefits to the community offers little to each participating individual’ (Rose, Reference Rose1981). It is becoming increasingly clear that interventions targeting physical health in people with psychosis are most effective when conducted within multidisciplinary settings with multi-modal targets and high intensity (McGinty et al. Reference McGinty, Baller, Azrin, Juliano-Bult and Daumit2015; Ward et al. Reference Ward, White and Druss2015). But would such interventions be feasible on a large scale, and if so, would they be sustainable long-term? Unfortunately, in most regions in the world, the likely answer is no to both of these questions. Previous research suggests that even with an effective intervention, the benefit is unlikely to be sustained after the cessation of the study (Alvarez-Jimenez et al. Reference Alvarez-Jimenez, Martinez-Garcia, Perez-Iglesias, Ramirez, Vazquez-Barquero and Crespo-Facorro2010; Gates et al. Reference Gates, Killackey, Phillips and Alvarez-Jimenez2015). Maintaining physical wellbeing is a life-long issue, and for the intervention to be meaningful, the benefit needs to be ongoing. In applying the concept of prevention paradox, perhaps we need to look for a ‘good enough’ intervention that is feasible and sustainable.

Epidemiology may be able to guide the way forward. We may be able to develop more theory-driven interventions to target mechanisms of change (Gates et al. Reference Gates, Killackey, Phillips and Alvarez-Jimenez2015) by using the behavioural epidemiology framework (Sallis et al. Reference Sallis, Owen and Fotheringham2000) to gain a better understanding of unique factors that influence the behaviour of interest in people with psychotic illness. Consider physical activity (PA), one of the core components of most behavioural interventions (Rosenbaum et al. Reference Rosenbaum, Tiedemann, Stanton, Parker, Waterreus, Curtis and Wardin press a ), as an example. PA is a complex behaviour. In order to understand why some people exercise and others do not, we need to appreciate the correlates and determinants of PA by utilising an epidemiological approach (Bauman et al. Reference Bauman, Reis, Sallis, Wells, Loos and Martin2012). In essence, correlates are the factors that show statistical associations with the behaviour identified through cross-sectional studies. Determinants, on the other hand, are the factors that show causal relationships with the behaviour which can only be identified through studies of aetiological design such as longitudinal studies or experimental studies (Bauman et al. Reference Bauman, Reis, Sallis, Wells, Loos and Martin2012). These factors exist at multiple levels ranging from the individual (e.g., type of antipsychotic medications people are taking), the regional policy (e.g., accessibility of recreational parks nearby), to the global environment (e.g., social and cultural attitude towards PA) (Bauman et al. Reference Bauman, Reis, Sallis, Wells, Loos and Martin2012). Gaining a better understanding of these factors will likely reveal useful insights into the essential components of effective interventions. While there is an increasing amount of evidence base for PA correlates in people with psychotic illness (Vancampfort et al. Reference Vancampfort, Knapen, Probst, Scheewe, Remans and De Hert2012, Reference Vancampfort, Correll, Probst, Sienaert, Wyckaert, De Herdt, Knapen, De Wachter and De Hert2013), there remains limited evidence on PA determinants based on longitudinal and experimental studies.

Once the ‘essential components’ of an effective PA, or any other behavioural intervention are identified through epidemiological studies, such evidence can be used to strengthen arguments for more funding to be put towards implementing population wide (e.g., targeted and specific lifestyle promotions), and service provision changes (e.g., making EPs and dieticians an essential part of mental health services) to reduce modifiable risk factors in people with psychotic illness. In other words, finding correlates and determinants may, at policy and global levels, aid us in changing the environment rather than the individuals to facilitate an improved lifestyle, thus bringing a large benefit to the people with psychotic illness as a population.

In addition, the significant correlates and determinants found at the individual level can aid us in finding helpful modifiable risk factors to improve our patients’ lifestyle. Each service can then build on these to create its own unique programme depending on the financial and resource constraints as well as its particular patient characteristics within the service. In fact, as KBIM suggests, such non-essential or site-specific components of the intervention may even hold the key to the success. Although not a psychiatric example, the Football Fans In Training study (Hunt et al. Reference Hunt, Wyke, Gray, Anderson, Brady, Bunn, Donnan, Fenwick, Grieve, Leishman, Miller, Mutrie, Rauchhaus, White and Treweek2014) showed that by making a lifestyle intervention to be relevant to the participants’ interest (i.e., football), nearly 90% of a traditionally difficult-to-engage population (obese middle-aged men) were able to complete a 12-month behavioural intervention programme with favourable outcomes. On the other hand, the implementation of smoke-free hospital policies in psychiatry wards, with little consideration to personal relevance for one of the most important modifiable risk factors for cardio-metabolic health, appears to have little effect on long-term smoking habits of the patients (Prochaska et al. Reference Prochaska, Fletcher, Hall and Hall2006). This is despite some evidence that a majority of people with psychotic illness want to quit smoking (Ashton et al. Reference Ashton, Rigby and Galletly2013), and they are able to do so when effective interventions are offered (Ashton et al. Reference Ashton, Rigby and Galletly2013; Stubbs et al. Reference Stubbs, Vancampfort, Bobes, De Hert and Mitchell2015). Therefore, such policies may, when combined with appropriate interventions that consider personal relevance to each patient, actually provide an ideal opportunity for mental health care providers to modify patients’ smoking habits.

Changing culture

One of the most important foundations for systemic change in how mental health services respond to the physical health of their patients lies in changing staff attitudes. For too long, mental health staff have focussed solely on improving patients’ mental wellbeing, while largely ignoring physical health issues. Re-establishing good physical health as a key treatment goal for mental health services is essential, with this goal being met by whatever means are available in a particular treatment service setting. We believe, as was the case 5 years ago (Crompton et al. Reference Crompton, Groves and McGrath2010), that if shared care models with primary care providers can be successfully implemented, this goal can be achieved. However, if this is not achievable, due to systemic difficulties in integrating primary care and mental health services, then mental health services will need to ensure that the physical health needs of their patients are met.

A key element for both scenarios is making mental health care providers aware of the importance of long-term physical health outcomes for their patients; obesity, hypertension, dyslipidaemia and sedentary behaviour may not be as readily apparent as extrapyramidal side effects or deliberate self-harm, yet they can seed long-term morbidity that will directly impact life expectancy. For example, this may mean favouring antipsychotic medications with less metabolic side effects and avoiding excess polypharmacy, in combination with appropriate non-pharmacological interventions (Gates et al. Reference Gates, Killackey, Phillips and Alvarez-Jimenez2015). Another strategy to achieve such awareness may be wellness programs that target mental health staff, demonstrating directly what PA and nutrition interventions entail, and developing a positive physical health culture that will drive change among mental health practitioners (Rosenbaum et al. Reference Rosenbaum, Watkins, Ward, Pearce, Fitzpatrick and Curtisin press b ). Such cultural change is likely to be essential in achieving sustainable outcomes (Gates et al. Reference Gates, Killackey, Phillips and Alvarez-Jimenez2015) – the ‘holy grail’ of translating short-term interventions into improved life expectancy in people with psychotic illness.

Preventing mortality from other causes

While there has been an emphasis on addressing mortality associated with cardiovascular risk factors, we must also address the gap in survival attributable to cancer in people living with psychotic illness. Even though the risk of developing cancer in people with psychotic illness is equivalent to that of the general population (Osborn et al. Reference Osborn, Limburg, Walters, Petersen, King, Green, Watson and Nazareth2013), the mortality arising from it is significantly higher (Kisely et al. Reference Kisely, Forsyth and Lawrencein press). In a population based New Zealand study (Cunningham et al. Reference Cunningham, Sarfati, Stanley, Peterson and Collings2015), people with psychotic illness were approximately two and a half times more likely to die within 5 years following diagnosis with breast cancer and three times more likely to die from colorectal cancer compared with those with no recent mental health service contact diagnosed with the same malignancies. Later stage of malignancy at the time of diagnosis explained more than one-third of the survival difference while the presence of other physical comorbidities in those with psychotic illness further contributed to reduced survival outcomes (Cunningham et al. Reference Cunningham, Sarfati, Stanley, Peterson and Collings2015). Even after adjusting for stage of cancer at diagnosis and comorbidities, there still remained a significant survival disadvantage for people with psychotic illness, possibly as a result of them being less likely to receive surgery or chemotherapy (Baillargeon et al. Reference Baillargeon, Kuo, Lin, Raji, Singh and Goodwin2011; Kisely et al. Reference Kisely, Crowe and Lawrence2013).

Congruent with lower rates of screening for cardiovascular risk factors, people with psychotic illness are less likely to be screened for breast, cervical, prostate and colorectal cancer (Happell et al. Reference Happell, Scott and Platania-Phung2012) contributing to diagnostic delays. For example, population-based retrospective studies of health records in Canada revealed that women with schizophrenia were approximately 30% less likely to have received cervical cancer screening (Martens et al. Reference Martens, Chochinov, Prior, Fransoo, Burland and Need To Know2009) and almost 40% less likely to have had mammography (Chochinov et al. Reference Chochinov, Martens, Prior, Fransoo, Burland and Need To Know2009). This further emphasises the importance of improved integration in service delivery between mental health, primary care and specialist oncology sectors to address the suboptimal screening and treatment of cancer.